Hi Class of 2011,
I finally sent in the letter grades to ASA today. The grade distribution remains roughly what it was last year - around 10% earned an A, 49% secured A-, about 31.7% obtained a B and the rest a B-.
Was glad to see some familiar names do well and was a little disappointed another set of familiar names do less well than I'd expected. By familiar names I mean people I have come to interact with and know in the course of the course.
So, again, all the best for the journey ahead. Do keep in touch going fwd and should you actually put into practice what you picked up in MKTR, I would love to hear about it. Should you as an alum tomorrow want to talk as a guest speaker to the MKTR class etc, that would be just great too.
Ciao and cheerios,
Sudhir
Tuesday, December 7, 2010
Sunday, December 5, 2010
Grading related mailbag
Two emails I received (made anonymous) and my responses therein. IMHO, bear wider dissemination:
On peer eval impact on project grading:
My response:
On phase II marks allocation:
My response:
On peer eval impact on project grading:
Can you please let us know the weight given on peer evaluation out of the total weight of 40 marks on the group project.
Regards
R
My response:
Its a maximum of 4 out of the 40 for the project.
And it is given only if there is concurrence (2 or more people in the group have peer rated a deviation similarly). We only had a handful of cases where the criteria were met.
Sudhir
On phase II marks allocation:
Apoorva,
First thing I said I am fine with the marks for phase III.
I just rechecked the standings on the number of responses which was put up on the professor’s blog finally.
We were placed X th out of 37 teams.
So, we were higher than median even without considering the ISB responses and also that we were a smaller team of 6.
Now, it was communicated earlier to us that we will be given credit only for non-ISB responses and also depending on the team size.
So, I do not understand why we should be getting a score of 7 which is below average. Clearly we performed better than more than ½ the teams so we would expect to get well above average.
And as the prof. has pointed himself in his blog post there is not a big variation in any component other than this we cannot afford to get less on this when we deserve credit.
Regards,
V
My response:
Let me answer this one.
1. ISB responses were considered except where IP addresses were clearly duplicated. The large numbers of foreign and otherwise recently nonresident students in several teams would have been disadvantaged otherwise. However, ISB responses were discouraged for others and rightly so. In any case we had about 10% of the total responses from within campus – not a magnitude that would reverse the current rank order or anything.
2. Groups that made the minimum threshold – 10 valid responses per team member – got the baseline score of 7. This is the vast majority of teams – 26 out of the 36 we had (Apoorva, correct me if I’m mistaken here).
3. The top 5 teams in the total responses sweepstakes got additional credit of 1.5. Their totals were far higher than average responses/team, the responses seemed genuine and regardless of whether or not one normalizes by team size, their output was high and thereby merited additional credit – so I deemed.
4. The bottom 5 teams got about 0.5 points less than the baseline 7, depending on the actuals.
5. Over all the average was 7.18 and you stand where you do.
Now, while I make no claims that this particular grading scheme was perfect or anything, it is consistent and objective. IMHO, of course. Changing anything at this stage will kickoff changes for groups above yours as well as those par –the vast majority of groups there are – and will likely nullify any grade related gains one may hope to make thereby. Hence, I shall go with what is currently done.
Sorry abt the delay in letter grades release. Shall be done soon.
Sudhir
Wednesday, December 1, 2010
We're putting the grades out soon
Hi all,
We'll be putting up the component grades and the overall letter grades out shortly.
A few quick points: (1.) There has been substantial revision of midterm answers - both in the MCQs and in the text portion. In particular, there's an entire 8 mark Q in part B that is being 'removed' from the exam. So everybody gets those 8 marks regardless of level of attempt. However, where folks have already made some attempt and received partial credit, the corresponding increase in their midterm tally will be lesser. Hence, NOTE: the increase may be uneven across students.
(2.) Regarding quizzes and attendance, the grades are concentrated along a narrow band - as expected. Folks who have missed a quiz or two will have their marks imputed on some pro rata basis. I did not intend to have the quiz/CP portion be some big source of variation to start with.
However, I must note with some concern that the attendance tallies are too high to be realistic. For sure, I saw people during the presentations who I don't ever recall seeing in class. No doubt, proxies have been all the rage. That is unfortunate. I did want the folks who did diligently attend class to get some credit - some 1% point odd more but now that is impossible to rectify at this stage.
(3) The project grading has also been along a rather narrow band for phases 1 and 2. Only the extreme outlier 5 teams in the whole have gotten extra credit (1.5% more) than the others in phase 2. Phase 3 grades have been normalized across sections - so that section averages are comparable. Not much normalization was needed, as it turned out.
(4). Managed to get through to getting the lucky draw done. Its someone who filled out for team Atri. She'll be getting sodexho coupons worth INR 5000/-.
Have asked the AAs to put up the class avg and std dev on blackboard.
Chalo, that's it for now.
Sudhir
We'll be putting up the component grades and the overall letter grades out shortly.
A few quick points: (1.) There has been substantial revision of midterm answers - both in the MCQs and in the text portion. In particular, there's an entire 8 mark Q in part B that is being 'removed' from the exam. So everybody gets those 8 marks regardless of level of attempt. However, where folks have already made some attempt and received partial credit, the corresponding increase in their midterm tally will be lesser. Hence, NOTE: the increase may be uneven across students.
(2.) Regarding quizzes and attendance, the grades are concentrated along a narrow band - as expected. Folks who have missed a quiz or two will have their marks imputed on some pro rata basis. I did not intend to have the quiz/CP portion be some big source of variation to start with.
However, I must note with some concern that the attendance tallies are too high to be realistic. For sure, I saw people during the presentations who I don't ever recall seeing in class. No doubt, proxies have been all the rage. That is unfortunate. I did want the folks who did diligently attend class to get some credit - some 1% point odd more but now that is impossible to rectify at this stage.
(3) The project grading has also been along a rather narrow band for phases 1 and 2. Only the extreme outlier 5 teams in the whole have gotten extra credit (1.5% more) than the others in phase 2. Phase 3 grades have been normalized across sections - so that section averages are comparable. Not much normalization was needed, as it turned out.
(4). Managed to get through to getting the lucky draw done. Its someone who filled out for team Atri. She'll be getting sodexho coupons worth INR 5000/-.
Have asked the AAs to put up the class avg and std dev on blackboard.
Chalo, that's it for now.
Sudhir
Monday, November 15, 2010
The desi savings story continues....
Check this out - opportune or what?
Indian household savings to be 15x in a decade
OK, even an avid proponent of breakthrough/non-linear growth curves and all like yours truly has difficulty believing some of those numbers there. The growth rate required for a 15x leap in household savings within a decade are beyond reach. By far. A 3x to 5x leap to ~ 1-1.5 trillion in 2010 USD is certainly do-able though. Sure, in nominal terms, if, say, the USD crashes by 50% or something, the figure quoted above may just about become possible. Who knows.
Which is why I say, be skeptical. It saves so much more time and trouble than otherwise. Wonder what kind of methodology the 'study' has employed here. I for one would certainly like to examine their data, assumptions and choices made.
Here's more meat directly from the Project scope....
Hope that interlude was useful.
Sudhir
Indian household savings to be 15x in a decade
NEW DELHI: India may takeover the US in terms of households savings by 2020, touching the USD 5 trillion mark, as the growing economy would boost incomes of people, a study has said. "The household savings rate of 33.4 per cent would translate into incremental savings of USD 5 trillion over the next decade with growing incomes of Indian households," Assocham President Swati A Piramal said in a statement. Currently, the size of Indian household savings is over USD 330 billion, according to a study jointly done by Assocham and PricehWaterhouse Coopers.
OK, even an avid proponent of breakthrough/non-linear growth curves and all like yours truly has difficulty believing some of those numbers there. The growth rate required for a 15x leap in household savings within a decade are beyond reach. By far. A 3x to 5x leap to ~ 1-1.5 trillion in 2010 USD is certainly do-able though. Sure, in nominal terms, if, say, the USD crashes by 50% or something, the figure quoted above may just about become possible. Who knows.
Which is why I say, be skeptical. It saves so much more time and trouble than otherwise. Wonder what kind of methodology the 'study' has employed here. I for one would certainly like to examine their data, assumptions and choices made.
Here's more meat directly from the Project scope....
Indian households have traditionally preferred safety of bank deposits and government saving schemes. Less than 10 per cent of their investments are in other financial assets like shares, debenture and mutual funds, which is low as compared to some of the developed economies, Piramal said. "Given the quantum of savings, the need to mobilise savings into productive channels and the opportunity for financial intermediation, the next decade will be an opportunity of a lifetime for Indian capital markets," the study said.Aha, eh?
Hope that interlude was useful.
Sudhir
Friday, November 12, 2010
Project presentations learnings summary.
Class,
As I write this, am done with all group presentations. Would rather pen down learnings early on before memory and other priorities take their toll.
First off, given the time and other constraints (you had less than a week between dataset-access and deliverable presentation), the output was by and large commendable. Some quick general observations:
1. Research Objective (R.O.) matters.
Recall from lectures 1, 2 & 3 my repeated exhortations that "A clear cut R.O. that starts with an action verb defined over a crisp actionable object sets the agenda for all that follows". Well that wasn't all blah-blah blah. Its effects are measurable, as I came to see.
Suppose the entire group was on board with and agreed upon a single, well-defined R.O., then planning, delegation and recombining different modules into a whole would have been much simplified. Coherence matters much in a project this complex and with coordination issues of the kind you must've faced. It was likely to visibly impact the quality of the outcome, and IMHO, it did.
2. Two broad approaches emerged - Asset First and Customer First.
One, where you define your research objective (R.O.) as "Identify the most attractive asset class." and the other, "Identify the most attractive customer segment." The two R.O.s lead to 2 very different downstream paths.
Most groups preferred the first (asset first) route. Here, the game was to ID the most attractive asset classes using size, monetary value as addressable market or some such criterion and then filter in only those respondents who showed some interest in the selected asset classes. Then characterize the indirect respondent segments obtained and build recommendations on that basis.
I was trying to nudge people towards the second, "Customer segmentation first" route partly because it aligns much more closely with the core Marketing STP (Segmentation-Targeting-Positioning) way. In this approach, the entire respondent base is first segmented along psychographic- behavioral - motivational or demographic bases, then different segments are evaluated for attractiveness based on some criterion - monetary value, count or share etc, and then the most attractive segments are profiled/analyzed for asset class preferences and investments.
Am happy to say that in a majority of the groups, once a group implicitly chose a particular R.O., the approach that followed was logically consistent with the R.O.
3. Some novel, surprising things.
Just reeling off a few quick ones that do come to my mind.
One, how do you select the "most attractive" segment or asset class given a set of options? Some groups went with a simple count criterion - count the # of respondents corresponding to that cluster and pick the largest one. Some groups went further and used a value criterion - multiply the count with (%savings times average income times % asset class allocation) to arrive at a rupee figure. This latter approach is more rigorous and objective, IMHO. There were only 2 groups that went even further in their choice of a attractiveness criterion - the customer lifetime value (CLV) criterion. They multiplied the rupee value per annum per respondent with a (cleaned up) "years to retirement" variable to obtain the revenue stream value of a respondent over his/her pre-retirement lifetime. Post-retirement, people become net consumers and not net savers, so post-retirement is a clean break from pre-retirement. I thought this last approach was simply brilliant. Wow. And even within the two groups that did use this idea, one went further and normalized cluster lifetime earnings by cluster size giving a crisp comparison benchmark.
Two, how to select the basis variables for a good clustering solution? Regardless of which approach you took, a good segmenting solution in which clusters are clear, distinct, sizeable and actionable would be required. One clear thing that emerged across multiple groups was that using only the Q27 psychographics and the Demographics wasn't yielding a good clustering solution. The very first few runs (<1 minute each on JMP and I'm told several minutes on MEXL) should have signaled that things were off with this approach. Adding more variables would have been key. Typically, groups adding savings motivation variables, Q7 constant sum etc were able to see a better clustering solution. There is seldom any ideal clustering solution and that's a valuable learning when dealing with real data (unlike the made-up data of classroom examples).
One group that stood out in the second point approach used all 113 variables in the dataset in a factor analysis -> got some 43 odd factors -> labeled and IDed them -> then selectively chose 40 from among the 43 as a segmenting basis and obtained a neat clustering solution. The reason this approach stands out in my mind 'brute force approach' is that there's no place for subjective judgment, no chance that some correlations among disparate variables will have been overlooked etc. It's also risky as such attempts are fraught with multi-collinearity and inference issues. Anyway, it seemed to have worked.
Three, when entry into real estate financial products, a regional break up that plays upon the geographic variation captured in the primary data would have been a good thing from the XYZ point of view. Would help them prioritize and better allocate resources in managerial decision making etc.
Anyway, like I have repeatedly mentioned - effort often correlates positively with learning. So I'm hoping your effort on this project did translate into enduring learning regarding research design, data work, modeling, project planning, delegation and coordination among other things.
I wish you all the best in your future endeavors.
Goodbye and Goodluck.
Sudhir
P.S.
Would appreciate constructive feedback that may help the MKTR course next year. Many thanks to the kindred souls in Section D for the 'Thank You' card, I thought that was sweet.
As I write this, am done with all group presentations. Would rather pen down learnings early on before memory and other priorities take their toll.
First off, given the time and other constraints (you had less than a week between dataset-access and deliverable presentation), the output was by and large commendable. Some quick general observations:
1. Research Objective (R.O.) matters.
Recall from lectures 1, 2 & 3 my repeated exhortations that "A clear cut R.O. that starts with an action verb defined over a crisp actionable object sets the agenda for all that follows". Well that wasn't all blah-blah blah. Its effects are measurable, as I came to see.
Suppose the entire group was on board with and agreed upon a single, well-defined R.O., then planning, delegation and recombining different modules into a whole would have been much simplified. Coherence matters much in a project this complex and with coordination issues of the kind you must've faced. It was likely to visibly impact the quality of the outcome, and IMHO, it did.
2. Two broad approaches emerged - Asset First and Customer First.
One, where you define your research objective (R.O.) as "Identify the most attractive asset class." and the other, "Identify the most attractive customer segment." The two R.O.s lead to 2 very different downstream paths.
Most groups preferred the first (asset first) route. Here, the game was to ID the most attractive asset classes using size, monetary value as addressable market or some such criterion and then filter in only those respondents who showed some interest in the selected asset classes. Then characterize the indirect respondent segments obtained and build recommendations on that basis.
I was trying to nudge people towards the second, "Customer segmentation first" route partly because it aligns much more closely with the core Marketing STP (Segmentation-Targeting-Positioning) way. In this approach, the entire respondent base is first segmented along psychographic- behavioral - motivational or demographic bases, then different segments are evaluated for attractiveness based on some criterion - monetary value, count or share etc, and then the most attractive segments are profiled/analyzed for asset class preferences and investments.
Am happy to say that in a majority of the groups, once a group implicitly chose a particular R.O., the approach that followed was logically consistent with the R.O.
3. Some novel, surprising things.
Just reeling off a few quick ones that do come to my mind.
One, how do you select the "most attractive" segment or asset class given a set of options? Some groups went with a simple count criterion - count the # of respondents corresponding to that cluster and pick the largest one. Some groups went further and used a value criterion - multiply the count with (%savings times average income times % asset class allocation) to arrive at a rupee figure. This latter approach is more rigorous and objective, IMHO. There were only 2 groups that went even further in their choice of a attractiveness criterion - the customer lifetime value (CLV) criterion. They multiplied the rupee value per annum per respondent with a (cleaned up) "years to retirement" variable to obtain the revenue stream value of a respondent over his/her pre-retirement lifetime. Post-retirement, people become net consumers and not net savers, so post-retirement is a clean break from pre-retirement. I thought this last approach was simply brilliant. Wow. And even within the two groups that did use this idea, one went further and normalized cluster lifetime earnings by cluster size giving a crisp comparison benchmark.
Two, how to select the basis variables for a good clustering solution? Regardless of which approach you took, a good segmenting solution in which clusters are clear, distinct, sizeable and actionable would be required. One clear thing that emerged across multiple groups was that using only the Q27 psychographics and the Demographics wasn't yielding a good clustering solution. The very first few runs (<1 minute each on JMP and I'm told several minutes on MEXL) should have signaled that things were off with this approach. Adding more variables would have been key. Typically, groups adding savings motivation variables, Q7 constant sum etc were able to see a better clustering solution. There is seldom any ideal clustering solution and that's a valuable learning when dealing with real data (unlike the made-up data of classroom examples).
One group that stood out in the second point approach used all 113 variables in the dataset in a factor analysis -> got some 43 odd factors -> labeled and IDed them -> then selectively chose 40 from among the 43 as a segmenting basis and obtained a neat clustering solution. The reason this approach stands out in my mind 'brute force approach' is that there's no place for subjective judgment, no chance that some correlations among disparate variables will have been overlooked etc. It's also risky as such attempts are fraught with multi-collinearity and inference issues. Anyway, it seemed to have worked.
Three, when entry into real estate financial products, a regional break up that plays upon the geographic variation captured in the primary data would have been a good thing from the XYZ point of view. Would help them prioritize and better allocate resources in managerial decision making etc.
Anyway, like I have repeatedly mentioned - effort often correlates positively with learning. So I'm hoping your effort on this project did translate into enduring learning regarding research design, data work, modeling, project planning, delegation and coordination among other things.
I wish you all the best in your future endeavors.
Goodbye and Goodluck.
Sudhir
P.S.
Would appreciate constructive feedback that may help the MKTR course next year. Many thanks to the kindred souls in Section D for the 'Thank You' card, I thought that was sweet.
Wednesday, November 10, 2010
For folks who have missed quizzes
Hi all,
There have been folks who have happened to miss quizzes for whatever reason. I'm writing this up to say, 'Don;t worry about it'. You won't get a zero in the quiz you've missed. We'll find some way to impute your quiz marks such that nobody - you or the rest of your section - is disadvantaged in any way.
Sudhir.
There have been folks who have happened to miss quizzes for whatever reason. I'm writing this up to say, 'Don;t worry about it'. You won't get a zero in the quiz you've missed. We'll find some way to impute your quiz marks such that nobody - you or the rest of your section - is disadvantaged in any way.
Sudhir.
Project presentations Update
Class,
Just reviewed the first 10 presentations. I must impressive, given the time crunch and other constraints you folks were operating under.
The remaining groups which are to present, here's some updates for you.
Am told there are some evening events groups do not want to miss. At the same time, I would like this thing over for sections A and B today. Hence, this is what I propose:
If at least half a group's members can be present, then they are welcome to make their project presentation at any convenient time slot of thier choice between 2 and 9.30 pm today in my office or in the AC2 conference room (where I'm trying to arrange for a projector and computer).
The same shall go for sections C and D tomorrow as well.
Sudhir
Just reviewed the first 10 presentations. I must impressive, given the time crunch and other constraints you folks were operating under.
The remaining groups which are to present, here's some updates for you.
Am told there are some evening events groups do not want to miss. At the same time, I would like this thing over for sections A and B today. Hence, this is what I propose:
If at least half a group's members can be present, then they are welcome to make their project presentation at any convenient time slot of thier choice between 2 and 9.30 pm today in my office or in the AC2 conference room (where I'm trying to arrange for a projector and computer).
The same shall go for sections C and D tomorrow as well.
Sudhir
Tuesday, November 9, 2010
Phase III: The way forward
Class,
The last of the traditional lectures for the last of the sections are done. The only thing remaining is the course project phase III. Some guidelines on what to expect and do:
1. In each section, the order in which groups will present will be randomly drawn. About 5, at most 6 group presentations can be accommodated in lec 10 class time itself. The other groups shall be given time slots to present between 7.30 and 9.30 pm at the AC2 LT/ MLT or NMLT on the same day as their lecture 10.
2. The time between the start of 2 consecutive presentations should be no more than 20 minutes. Ideally, budget 15 minutes for the presentation and about 3-4 minutes for Q&A. Short clarification Questions can be asked anytime during the presentation but the more long-winded questions should come only at the end. Each time will be shown a 2 minute warning sign before their 15 minutes are up.
3. The format in which the submission files are named and the content format is explained briefly in this blog post here. Kindly adhere to the same as far as possible. Use of slide animations, highlighting, multimedia etc to better make a point are welcome.
4. Only the PPTs that have been submitted prior to the blackboard dropbox deadline can be used for the presentations. We will bring the PPTs and load them a priori onto the Classroom computer. No laptop from your side would be required for the presentation.
5. Kindly ensure your entire group is present for the presentation and is reasonably up to date on and on-board with broad aspects of the project.
6. Some basic common-sensical ground rules - (i) Kindly refrain from entering or leaving the classroom when a presentation is in progress. It is needlessly distracting. Kindly use the interval between two presentations to exit/enter. (ii) Kindly feel free to ask questions from a group but no more than 1 question from 1 student/group to a presenting group.
7. Please sign and hand over the peer evaluation form to your section AAs anytime between Wednesday to Friday.
8. The evlauation criteria for Phase III include, broadly -
(A) The quality and clarity of the methodology
(B) Logical flow and consistency of Analysis
(C) Clarity of communication
(D) Handling of Q&A
(E) Quality of recommendations
Pls feel free to use whatever software works for you - MEXL, SAS, SPSS whatever.
Hope that clarifies.
Sudhir
The last of the traditional lectures for the last of the sections are done. The only thing remaining is the course project phase III. Some guidelines on what to expect and do:
1. In each section, the order in which groups will present will be randomly drawn. About 5, at most 6 group presentations can be accommodated in lec 10 class time itself. The other groups shall be given time slots to present between 7.30 and 9.30 pm at the AC2 LT/ MLT or NMLT on the same day as their lecture 10.
2. The time between the start of 2 consecutive presentations should be no more than 20 minutes. Ideally, budget 15 minutes for the presentation and about 3-4 minutes for Q&A. Short clarification Questions can be asked anytime during the presentation but the more long-winded questions should come only at the end. Each time will be shown a 2 minute warning sign before their 15 minutes are up.
3. The format in which the submission files are named and the content format is explained briefly in this blog post here. Kindly adhere to the same as far as possible. Use of slide animations, highlighting, multimedia etc to better make a point are welcome.
4. Only the PPTs that have been submitted prior to the blackboard dropbox deadline can be used for the presentations. We will bring the PPTs and load them a priori onto the Classroom computer. No laptop from your side would be required for the presentation.
5. Kindly ensure your entire group is present for the presentation and is reasonably up to date on and on-board with broad aspects of the project.
6. Some basic common-sensical ground rules - (i) Kindly refrain from entering or leaving the classroom when a presentation is in progress. It is needlessly distracting. Kindly use the interval between two presentations to exit/enter. (ii) Kindly feel free to ask questions from a group but no more than 1 question from 1 student/group to a presenting group.
7. Please sign and hand over the peer evaluation form to your section AAs anytime between Wednesday to Friday.
8. The evlauation criteria for Phase III include, broadly -
(A) The quality and clarity of the methodology
(B) Logical flow and consistency of Analysis
(C) Clarity of communication
(D) Handling of Q&A
(E) Quality of recommendations
Pls feel free to use whatever software works for you - MEXL, SAS, SPSS whatever.
Hope that clarifies.
Sudhir
Monday, November 8, 2010
Project related travails
Hi All,
I've been meeting groups all afternoon and evening today and some things have come up which IMHOmerit wider dissemination:
1. Have some basic roadmap in mind before you start: This is important else you risk getting lost in the data and all the analyses that are now possible. There are literally millions of ways in which a dataset that size can be sliced and diced. Groups that had no broad, big-picture idea of where they want to go with the analysis inevitably run into problems.
Now don't get me wrong, this is not to pre-judge or straitjacket your perspective or anything - the initial plan you have in mind doesn't restrict your options. It can and should be changed and improvised as the analysis proceeds.
Update: OK. Some may ask - can we get a more specific example? Here is what I had in mind when I was thinking broad, basic plan from an example I outlined in the comments section to a post below:
2. Segmentation is the key: The Project essentially, at its core, boils down to an STP or Segmentation-Targeting-Positioning exercise. And it is the Segmentation part which is crucial to getting the TP parts right. What inputs to have for the segmentation part, what clustering bases to use, how many clusters to get out via k-means, how best to characterize those clusters and how to decide which among them is best/most attractive are, IMHO, the real tricky questions in the project.
3. Kindly ask around for JMP gyan: A good number of folk I have met seemed to have basic confusion regarding factor and cluster analyses and how to run these on JMP. This after I thought I'd done a good job going step-by-step over the procedure in class and interpreting the results. Kindly ask around for clarifications etc on the JMP implementation of these procedures. The textbook contains good overviews about the conceptual aspects of these methods.
I'm hopeful that at least a few folk in each group have a handle on these critical procedures - factor and cluster. I shall, for completeness sake, again go through them quickly tomorrow in class.
4. The 80-20 rule applies very much so in data cleaning: Chances are under 20% of the columns in the dataset will yield over 80% of its usable information content. So don't waste time cleaning data (i.e. removing missing values, nonsense answers etc) from all the columns, just the important ones only. Again, you need to have some basic plan in mind before you can ID the important columns.
Also, not all data cleaning need mean dropping rows. In some instances, missing values can perhaps be safely imputed using column means or medians or the mode (depending on data type).
Chalo, enough for now. More as updates occur.
Sudhir
I've been meeting groups all afternoon and evening today and some things have come up which IMHOmerit wider dissemination:
1. Have some basic roadmap in mind before you start: This is important else you risk getting lost in the data and all the analyses that are now possible. There are literally millions of ways in which a dataset that size can be sliced and diced. Groups that had no broad, big-picture idea of where they want to go with the analysis inevitably run into problems.
Now don't get me wrong, this is not to pre-judge or straitjacket your perspective or anything - the initial plan you have in mind doesn't restrict your options. It can and should be changed and improvised as the analysis proceeds.
Update: OK. Some may ask - can we get a more specific example? Here is what I had in mind when I was thinking broad, basic plan from an example I outlined in the comments section to a post below:
E.g. - First we clean data out for missing values in Qs 7,10,27 etc -> then do factor analysis on psychogr and demogr -> then did cluster analysis on the factors -> then we estimate segment sizes thus obtained -> then we look up supply side options -> arrive at recommendations.
Hope that clarifies.
2. Segmentation is the key: The Project essentially, at its core, boils down to an STP or Segmentation-Targeting-Positioning exercise. And it is the Segmentation part which is crucial to getting the TP parts right. What inputs to have for the segmentation part, what clustering bases to use, how many clusters to get out via k-means, how best to characterize those clusters and how to decide which among them is best/most attractive are, IMHO, the real tricky questions in the project.
3. Kindly ask around for JMP gyan: A good number of folk I have met seemed to have basic confusion regarding factor and cluster analyses and how to run these on JMP. This after I thought I'd done a good job going step-by-step over the procedure in class and interpreting the results. Kindly ask around for clarifications etc on the JMP implementation of these procedures. The textbook contains good overviews about the conceptual aspects of these methods.
I'm hopeful that at least a few folk in each group have a handle on these critical procedures - factor and cluster. I shall, for completeness sake, again go through them quickly tomorrow in class.
4. The 80-20 rule applies very much so in data cleaning: Chances are under 20% of the columns in the dataset will yield over 80% of its usable information content. So don't waste time cleaning data (i.e. removing missing values, nonsense answers etc) from all the columns, just the important ones only. Again, you need to have some basic plan in mind before you can ID the important columns.
Also, not all data cleaning need mean dropping rows. In some instances, missing values can perhaps be safely imputed using column means or medians or the mode (depending on data type).
Chalo, enough for now. More as updates occur.
Sudhir
Some project related Qs and clarifications
Folks,
Based on queries and comments received in class today re the project, let me try to quickly putup stuff that IMHO mertits wider dissemination.
1. Let me rush to clarify that no great detail is expected in the supply side Q.
As in, you're not expected to say - "XYZ should offer an FD [fixed deposit] with a two year minimum lock-in offering 8.75% p.a.".
No.
Saying "XYZ should offer FDs in its product portfolio." is sufficient.
2. Make the assumption that the sample adequately represents the target population - of young, urban, upwardly mobile professionals.
3. Yes, data cleaning is a long, messy process. But it is worthwhile since once it's done, the rest of the analyses follow through very easily indeed, in seconds.
4. It helps to start with some idea or set of conjectures about a set of product classes and a set of potential target segments in mind, perhaps. One can then use statistical analyses to either confirm or disprove particular hypotheses about them.
5. There is no 'right or wrong' approach to the problem. There is however a logical, coherent and data-driven approach to making actionable recommendations versus one that is not. I'll be looking for logical errors, coherency issues, unsustainable assumptions and the like in your journey to the recommendations you make in phase III.
Hope that clarifies.
Sudhir
Based on queries and comments received in class today re the project, let me try to quickly putup stuff that IMHO mertits wider dissemination.
1. Let me rush to clarify that no great detail is expected in the supply side Q.
As in, you're not expected to say - "XYZ should offer an FD [fixed deposit] with a two year minimum lock-in offering 8.75% p.a.".
No.
Saying "XYZ should offer FDs in its product portfolio." is sufficient.
2. Make the assumption that the sample adequately represents the target population - of young, urban, upwardly mobile professionals.
3. Yes, data cleaning is a long, messy process. But it is worthwhile since once it's done, the rest of the analyses follow through very easily indeed, in seconds.
4. It helps to start with some idea or set of conjectures about a set of product classes and a set of potential target segments in mind, perhaps. One can then use statistical analyses to either confirm or disprove particular hypotheses about them.
5. There is no 'right or wrong' approach to the problem. There is however a logical, coherent and data-driven approach to making actionable recommendations versus one that is not. I'll be looking for logical errors, coherency issues, unsustainable assumptions and the like in your journey to the recommendations you make in phase III.
Hope that clarifies.
Sudhir
Sunday, November 7, 2010
Phase III deliverables
Class,
Your deliverable consists of a 15 + 3 minute presentation of a 30 slide PPT (excluding appendices).
Deadline for dropbox submission of PPT is morning 0600 hrs on 10-Nov (Wednesday).
You are expected to answer two questions, primarily:
1. What products/Services should XYZ offer?
2. What customer segment should be targetted with (1)?
'Why' is a corollary to both questions above. Be prepared to make and argue your case effectively using evidence from primary and/or secondary data.
More specifically, your PPT should contain:
1. Filename (mandatory) - should be groupname_section.pptx when submitted into the dropbox on blackboard.
2. Title slide (mandatory) - project title, Group Name, member names and UIDs & MKTR section.
3. Presentation Outline (Optional) - a Contents page that outlines your presentation along sections (e.g. methodology, Q1, Q2, recommendations, etc).
4. A Methodology Section (Mandatory) - In preferably a graph or flowchart form, lay down what analysis procedures you used in what order to answer Q2 to reach the recommendations based on Q2.
5. A Data section (Mandatory) - explains that nature and structure of the data that were used. Be very brief but very informative - write (ii) the dimensions of the data matrices used as input to in different procedures, and (ii) the sources of data - cited sources if secondary data are used, and Question numbers in the survey questionnaire if primary data are used.
I strongly suggest using a tabular format here. Packs a lot of info into compact space. Easy to read and compare too.
6. Model Expressions (Mandatory)- Write the conceptual and/or mathematical expressions of any models used. Kindly place in the appendix section a descriptives table of the input data, a brief explanation of the X variables used, and of course, output tables along with interpretation.
7. Appendix Section (Optional): Some of the less important tables can be plugged into a separate appendix section (outside the 30 slide limit) in case you are running out of slide space. Have only the most important results tables in the main portion.
8. Recommendations (Mandatory) - crisp, clear, in simple words directed towards the client. Emphasize the usability and actionability of the recommendations.
Hope that clarifies.
Sudhir
Your deliverable consists of a 15 + 3 minute presentation of a 30 slide PPT (excluding appendices).
Deadline for dropbox submission of PPT is morning 0600 hrs on 10-Nov (Wednesday).
You are expected to answer two questions, primarily:
1. What products/Services should XYZ offer?
2. What customer segment should be targetted with (1)?
'Why' is a corollary to both questions above. Be prepared to make and argue your case effectively using evidence from primary and/or secondary data.
More specifically, your PPT should contain:
1. Filename (mandatory) - should be groupname_section.pptx when submitted into the dropbox on blackboard.
2. Title slide (mandatory) - project title, Group Name, member names and UIDs & MKTR section.
3. Presentation Outline (Optional) - a Contents page that outlines your presentation along sections (e.g. methodology, Q1, Q2, recommendations, etc).
4. A Methodology Section (Mandatory) - In preferably a graph or flowchart form, lay down what analysis procedures you used in what order to answer Q2 to reach the recommendations based on Q2.
5. A Data section (Mandatory) - explains that nature and structure of the data that were used. Be very brief but very informative - write (ii) the dimensions of the data matrices used as input to in different procedures, and (ii) the sources of data - cited sources if secondary data are used, and Question numbers in the survey questionnaire if primary data are used.
I strongly suggest using a tabular format here. Packs a lot of info into compact space. Easy to read and compare too.
6. Model Expressions (Mandatory)- Write the conceptual and/or mathematical expressions of any models used. Kindly place in the appendix section a descriptives table of the input data, a brief explanation of the X variables used, and of course, output tables along with interpretation.
7. Appendix Section (Optional): Some of the less important tables can be plugged into a separate appendix section (outside the 30 slide limit) in case you are running out of slide space. Have only the most important results tables in the main portion.
8. Recommendations (Mandatory) - crisp, clear, in simple words directed towards the client. Emphasize the usability and actionability of the recommendations.
Hope that clarifies.
Sudhir
Tying in Course learnings with phase III
Hi all,
Update: It seems to me that Q33 - media habits w.r.t. TV channels has been lost to us. There's *NO* variation in the responses anywhere in Q33 suggesting that the Q didn't work as planned. The ranking question template in qualtrics was pointlessly complex and involved some drag and drop operation that took me a while to figure. Respondents, exhausted at the fag end of the survey, can hardly be expected to get it right.
Well, now that we are dealing with REAL data, some of our learnings will be from such unfortunate events, I guess. Basically, we have lost all usable info on the media habits question. Will likely have to rely only on psychographics mostly to design a marketing communications message for the target segment now, I guess.
Continuing my last post:
We broke our research problem down into two parts - one relating to the supply side and the other, to the demand side. The supply side essentially asks what set of products the firm should offer and the demand side complements it by asking what customers want in financial/retirement planning products. The demand and supply sides are complementary and simultaneous and must be solved iteratively to arrive at a coherent, consistent answer.
Now that we have in some sense mapped the possible set of supply side options to a manageable set of target customer segments on the demand side, the Q arises, what specific analysis methods and techniques might come in use. Pls understand there's plenty of leeway here and many different ways to peel an apple. What I'll talk about here is merely 1 possible way among many alternatives.
Clearly, Factor and Cluster analyses can easily be brought into play - to reduce the # of variables in the psychographics for instance and to explore 'natural groupings' among customers on the basis of demographics (life-cycle stage.age or employment or family size etc), psychographics (risk-return appetite), asset class preference, something else altogether or some combination of all of these. Just remember Kotler gyan about what segments should be like - measurable, actionable, reachable, distinct and all that.
What about the ANOVA and regression stuff we learned? Specific hypotheses you have - such as e.g. "People in their 20s and early 30s are markedly more risk averse than those in their late 30s and 40s", or "there is systematic association between employment type and preference for fixed income/annuities post retirement" that feed into your storyline can and should be inferred using statistical analysis.
OK, what about discrete choice models (like the Logit) that we will do in lec 9? Well, turns out that just like ANOVA and multivariate regression use the 'Analyze > Fit Model' sequence in JMP, so does Logit as well. Hallelujah. Just input a categorical Y into the Y area and see, the 'Standard least squares' method type at the top-right of the screen changes to 'logistic analysis' automatically. Logistic analyses can be used to infer whether particular categorical Ys relate systematically to some set of Xs in a statistically significant way.
For instance, I could have a hypothesis saying "Public sector bank as primary bank choice relates systematically to age (older people perefer this), employment type (govt employees prefer this) and low risk-low return seekers" or something like that.
Well, applying techniques learned in class would be good and should be done wherever opportunity arises. Using methods learned outside of class is very welcome too, just be prepared for Qs on them.
The primary learning in the project comes from phase III, IMHO. It teaches us how to do phase I better based on the mistakes made in phase II questionnaire as reflected in the actual data collected. The biggest learning IMHO is a resetting of our own pre-conceived notions of what analyzing real data is like, of what can and cannot be reliably inferred from the data we do have and so on.
Well, that's it from me.
P.S.
Am prepping lec 9. Hope is to cover Discrete choice and a bit of MDS (Multidimensional scaling) under perceptual maps. Now am unable to locate good MDS tools in JMP and might have to fallback on R for this. Now, I see little scope for MDS usage in phase III, so a demo on R shouldn't be too troublesome, hopefully, IMHO.
Sudhir
Update: It seems to me that Q33 - media habits w.r.t. TV channels has been lost to us. There's *NO* variation in the responses anywhere in Q33 suggesting that the Q didn't work as planned. The ranking question template in qualtrics was pointlessly complex and involved some drag and drop operation that took me a while to figure. Respondents, exhausted at the fag end of the survey, can hardly be expected to get it right.
Well, now that we are dealing with REAL data, some of our learnings will be from such unfortunate events, I guess. Basically, we have lost all usable info on the media habits question. Will likely have to rely only on psychographics mostly to design a marketing communications message for the target segment now, I guess.
Continuing my last post:
We broke our research problem down into two parts - one relating to the supply side and the other, to the demand side. The supply side essentially asks what set of products the firm should offer and the demand side complements it by asking what customers want in financial/retirement planning products. The demand and supply sides are complementary and simultaneous and must be solved iteratively to arrive at a coherent, consistent answer.
Now that we have in some sense mapped the possible set of supply side options to a manageable set of target customer segments on the demand side, the Q arises, what specific analysis methods and techniques might come in use. Pls understand there's plenty of leeway here and many different ways to peel an apple. What I'll talk about here is merely 1 possible way among many alternatives.
Clearly, Factor and Cluster analyses can easily be brought into play - to reduce the # of variables in the psychographics for instance and to explore 'natural groupings' among customers on the basis of demographics (life-cycle stage.age or employment or family size etc), psychographics (risk-return appetite), asset class preference, something else altogether or some combination of all of these. Just remember Kotler gyan about what segments should be like - measurable, actionable, reachable, distinct and all that.
What about the ANOVA and regression stuff we learned? Specific hypotheses you have - such as e.g. "People in their 20s and early 30s are markedly more risk averse than those in their late 30s and 40s", or "there is systematic association between employment type and preference for fixed income/annuities post retirement" that feed into your storyline can and should be inferred using statistical analysis.
OK, what about discrete choice models (like the Logit) that we will do in lec 9? Well, turns out that just like ANOVA and multivariate regression use the 'Analyze > Fit Model' sequence in JMP, so does Logit as well. Hallelujah. Just input a categorical Y into the Y area and see, the 'Standard least squares' method type at the top-right of the screen changes to 'logistic analysis' automatically. Logistic analyses can be used to infer whether particular categorical Ys relate systematically to some set of Xs in a statistically significant way.
For instance, I could have a hypothesis saying "Public sector bank as primary bank choice relates systematically to age (older people perefer this), employment type (govt employees prefer this) and low risk-low return seekers" or something like that.
Well, applying techniques learned in class would be good and should be done wherever opportunity arises. Using methods learned outside of class is very welcome too, just be prepared for Qs on them.
The primary learning in the project comes from phase III, IMHO. It teaches us how to do phase I better based on the mistakes made in phase II questionnaire as reflected in the actual data collected. The biggest learning IMHO is a resetting of our own pre-conceived notions of what analyzing real data is like, of what can and cannot be reliably inferred from the data we do have and so on.
Well, that's it from me.
P.S.
Am prepping lec 9. Hope is to cover Discrete choice and a bit of MDS (Multidimensional scaling) under perceptual maps. Now am unable to locate good MDS tools in JMP and might have to fallback on R for this. Now, I see little scope for MDS usage in phase III, so a demo on R shouldn't be too troublesome, hopefully, IMHO.
Sudhir
Saturday, November 6, 2010
Some specific phase III gyan
Hi all,
Last year's experience tells me it's better to provide some (optional) guidance and structure than none at all, given the time and other constraints we are under. So here goes:
First, if you haven't already, then kindly first read these posts for context and continuation: General gyan on the project scope and background and Some phase I guidance.
There are broadly three research questions I can see relating (i) to a market entry decision, and subsequently (ii) to an STP or Segmentation-Targeting-Positioning decision for client firm XYZ. The first, or the zeroth question rather, is a preliminary one that must be acknowledged and gotten out of the way, first thing.
0. Is there a market out there at all?
1. If yes, what are the characteristics of this market?
2. How best to reach the attractive customer segments?
The zeroth research question implicitly asks if there exists an addressable, sizeable, profitable, sustainable market in financial services and/or products given the target population (the young, upwardly mobile, urban middle class).
Now, one may say, "well, look, there are so many firms and products in this space. Ergo, there must be a good, profitable, actionable market out there. No?" Well, likely but not necessarily. The last 2 years in particular teach us that in bubble scenarios, the herd is grossly wrong. We'll have to take this call for ourselves by ourselves. In this case though, I doubt you'll have much trouble making a case for entry versus non-entry. Hint: Broadly glance at what kind of numbers we are talking about here, at the macro-level - the equivalent of a full one-third of India's GDP, from the household sector alone, needs to find a home in some combination of asset classes, every year. Would be nice if (but is by no means mandatory that) your phase III analysis incorporate some thought, rationale and recommendation regarding the entry or non-entry for XYZ in the target space.
Now consider research Qs 1 and 2. Those simple looking research questions are actually quite involved. Q1 has two sides to it - the supply side characteristics of the market and those of the demand side. Let me label these sub-questions 1.1. and 1.2. thus:
1.1. (Supply side) What combination of products/services/solutions in what asset classes should XYZ consider offering?
1.2. (Demand side) What is the distribution of target population preferences over asset classes and over the attributes of financial products/services/solutions?
For the supply side, ask if there exist synergies in introducing a set of products rather than just one - in terms of competencies required, talent/skills available, scale, marketing, etc. Imaginative use and application of Secondary and Primary data would be great here. For instance, primary data can at least provide directly the (stated) distribution of current and prospective investments into broadly defined asset classes. Again, this is entirely my idea of how I would think if I were in your seat. It nowhere says I am 100% right in what I am thinking. Hence, you don't have to follow my lead here and can go with some radical, innovative and totally out-of-the-box approach of your own.
For the demand side, primary data in some measure allows us to collect data on target population preferences over financial product attributes (broadly speaking) such as risk/return, time to maturity, tax saving status, benefits sought, flexibility or liquidity at short notice, etc. This rich dataset also provides opportunities for Segmentation of the customer base, along various dimensions of interest (entirely your call what you want to go with). I would rather you try to go with what can be actionable from the client firm's viewpoint.
Once some sort of segmentation base has been decided upon, segmentation done and the segments identified, labeled, sized up and analyzed, then comes Q2 with a targetting and positioning plan for the most attractive segments identified. We've asked some basic Qs about media habits and psychographics etc to help with a rudimentary marketing plan. Essentially, you are able to go one step ahead and tell the client not just what are the most attractive customer segments in the market for so-and-so asset classes, but also how to reach them, communicate with them and sell them the product.
Now the logical Q arises, how did all that we do in class come in handy here? Aha. Good Q. I shall putup a separate post explaining and connecting some of the techniques we covered with some of the research Qs above.
OK, that is it from me for this part. I'd ideally you mull over this with your group. Again, you don't have to go with this plan outline as above but kindly ensure that you bring value to client XYZ.
I'd be happy to get some feedback, comments, queries and clarifications on Phase III thus far. Pls use the comments section below should any such arise.
Some of you may have already started playing the data. Please do. You'll find the practical difficulties associated with collating, filtering, sorting out, analyzing and interpreting real data. That, IMVHO is tremendous learning for the individual, the team and the class as a whole.
Sudhir
Last year's experience tells me it's better to provide some (optional) guidance and structure than none at all, given the time and other constraints we are under. So here goes:
First, if you haven't already, then kindly first read these posts for context and continuation: General gyan on the project scope and background and Some phase I guidance.
There are broadly three research questions I can see relating (i) to a market entry decision, and subsequently (ii) to an STP or Segmentation-Targeting-Positioning decision for client firm XYZ. The first, or the zeroth question rather, is a preliminary one that must be acknowledged and gotten out of the way, first thing.
0. Is there a market out there at all?
1. If yes, what are the characteristics of this market?
2. How best to reach the attractive customer segments?
The zeroth research question implicitly asks if there exists an addressable, sizeable, profitable, sustainable market in financial services and/or products given the target population (the young, upwardly mobile, urban middle class).
Now, one may say, "well, look, there are so many firms and products in this space. Ergo, there must be a good, profitable, actionable market out there. No?" Well, likely but not necessarily. The last 2 years in particular teach us that in bubble scenarios, the herd is grossly wrong. We'll have to take this call for ourselves by ourselves. In this case though, I doubt you'll have much trouble making a case for entry versus non-entry. Hint: Broadly glance at what kind of numbers we are talking about here, at the macro-level - the equivalent of a full one-third of India's GDP, from the household sector alone, needs to find a home in some combination of asset classes, every year. Would be nice if (but is by no means mandatory that) your phase III analysis incorporate some thought, rationale and recommendation regarding the entry or non-entry for XYZ in the target space.
Now consider research Qs 1 and 2. Those simple looking research questions are actually quite involved. Q1 has two sides to it - the supply side characteristics of the market and those of the demand side. Let me label these sub-questions 1.1. and 1.2. thus:
1.1. (Supply side) What combination of products/services/solutions in what asset classes should XYZ consider offering?
1.2. (Demand side) What is the distribution of target population preferences over asset classes and over the attributes of financial products/services/solutions?
For the supply side, ask if there exist synergies in introducing a set of products rather than just one - in terms of competencies required, talent/skills available, scale, marketing, etc. Imaginative use and application of Secondary and Primary data would be great here. For instance, primary data can at least provide directly the (stated) distribution of current and prospective investments into broadly defined asset classes. Again, this is entirely my idea of how I would think if I were in your seat. It nowhere says I am 100% right in what I am thinking. Hence, you don't have to follow my lead here and can go with some radical, innovative and totally out-of-the-box approach of your own.
For the demand side, primary data in some measure allows us to collect data on target population preferences over financial product attributes (broadly speaking) such as risk/return, time to maturity, tax saving status, benefits sought, flexibility or liquidity at short notice, etc. This rich dataset also provides opportunities for Segmentation of the customer base, along various dimensions of interest (entirely your call what you want to go with). I would rather you try to go with what can be actionable from the client firm's viewpoint.
Once some sort of segmentation base has been decided upon, segmentation done and the segments identified, labeled, sized up and analyzed, then comes Q2 with a targetting and positioning plan for the most attractive segments identified. We've asked some basic Qs about media habits and psychographics etc to help with a rudimentary marketing plan. Essentially, you are able to go one step ahead and tell the client not just what are the most attractive customer segments in the market for so-and-so asset classes, but also how to reach them, communicate with them and sell them the product.
Now the logical Q arises, how did all that we do in class come in handy here? Aha. Good Q. I shall putup a separate post explaining and connecting some of the techniques we covered with some of the research Qs above.
OK, that is it from me for this part. I'd ideally you mull over this with your group. Again, you don't have to go with this plan outline as above but kindly ensure that you bring value to client XYZ.
I'd be happy to get some feedback, comments, queries and clarifications on Phase III thus far. Pls use the comments section below should any such arise.
Some of you may have already started playing the data. Please do. You'll find the practical difficulties associated with collating, filtering, sorting out, analyzing and interpreting real data. That, IMVHO is tremendous learning for the individual, the team and the class as a whole.
Sudhir
Friday, November 5, 2010
General gyan on the project scope and background
Hi all,
One (foreign) student asked me what the logic was behind the survey questionnaire. And it is not surprising why some may say this. Knowing where one is coming from during the design phase (of the questionnaire, in this case) helps in charting an analysis course better, perhaps. I promised to make a writeup about the logic and the story behind the questionnaire. So, this is the story, the thinking and the background to the project scope.
We live in what the Chinese would call 'interesting times'. The economic crisis that manifested as a global financial markets meltdown continues to cast its shadow not just on current employment and home values but also on longer term issues such as the ability of governments and private organizations to live upto their pension and healthcare liabilities/ obligations. Regardless of whether we choose to acknowledge it or not, the fact remains that a lot of our old assumptions and beliefs about how the world works, including the bedrock currency arrangement coded in the Bretton woods II system, may no longer hold anymore.
Sample these articles -
The end of retirement (from The Economist, June 2010)
Jobs, pensions and markets (from The Economist, Aug 2010)
This has huge implications for us ordinary people as well, even in countries like India which have (so far) been less affected by the crisis.
Unlike Indian government employees, whose retirement and lifetime pension is guaranteed by the Government of India - which because it prints its own money is technically safe from bankruptcy and default - the rest of us cannot really say our retirement and pensions are guaranteed in any real measure. Many young people in the US for instance question why they have to pay into social security when chances are they will never get to see a penny from there by the time they retire. We are indeed in the vanguard of the generations that will have to plan for their futures in exacting detail in the backdrop of increasing uncertainty in the years to come.
That automatically brings up its own host of interesting questions - How will we do? What will do? More generally, what is the current crop of professionals in their 20s and 30s thinking and planning in this scenario? How much confidence do they have in the current set of assets and institutions to protect their savings against inflation? What are their concerns, their preferences and their constraints? How and how much does their view and their actual investment pattern depend on their age, gender, wealth, life cycle stage, nature of employment or other factors?
Mind you, this is not idle chatter - In the aggregate, Indian household savings have averaged about 32-35% of GDP annually in recent years. Unprecedented, really. And thanks to all the visionaries starting with Shri PVN Rao and onto Shri AB Vajpayee and then Shri Manmohan Singh for enabling this. Coupled with an Incremental Capital Output Ratio (ICOR) of about 4 to 1, we are essentially assured growth of about 8% p.a. in real terms for at least the next few years based on sustainable internal demand alone. The choices that our target population segment - the young upwardly mobile urban middle class - will make will be multiplied several times fold as these people tend to be opinion leaders and trend setters also. Where will this mass of money flow to - which asset classes? In what proportion? With what confidence should conditions change down the line? So many questions! The project is one attempt to bring some structure and order to this confusing mass of questions by the application of the scientific method and statistical analysis to understand this conundrum. OK, that is the big picture view of a very complex issue. You've already read the scope document, so you now also have a good handle on the actual problem XYZ corp is handling.
Now, kindly read through this post made here in full if you haven't already:
http://marketing-yogi.blogspot.com/2010/10/some-phase-i-guidance.html
This summarizes my initial views on how to tackle a question as complex as this - basically that:
(i) we limit the scope to what can be done realistically. In this case, we focus only on the big-picture, only at the broad asset class level and not go into details of asset sub-classes and particular schemes available therein. I chose to focus broadly only on Equities (including MFs, SIPs and demat equities), Fixed income plans (bonds, insurance policies that fall under this purview, certain pension and annuity plans etc), property (different types and in different types of cities)
(ii) we focus only on the distribution of asset class preferences in the target population - perhaps as functions of demographics and psychographics.
(iii) we come up with recommendations for which asset classes to go with - some sort of ranking perhaps - given the preference structure of the target population.Any secondary analysis that uses population sizes for such segments (e.g. from the Marketing Whitebook) would be a great plus, IMHO.
OK. This post is at a general, big-picture level and is long enough as-is. Shall continue with more specific guidance in subsequent posts.
Again, not all of you may find general gyan of this sort immediately useful. That is fine. Feel free to go with whatever you have in mind. However, have someone in the group look for updates on the specific guidance.
Sudhir
One (foreign) student asked me what the logic was behind the survey questionnaire. And it is not surprising why some may say this. Knowing where one is coming from during the design phase (of the questionnaire, in this case) helps in charting an analysis course better, perhaps. I promised to make a writeup about the logic and the story behind the questionnaire. So, this is the story, the thinking and the background to the project scope.
We live in what the Chinese would call 'interesting times'. The economic crisis that manifested as a global financial markets meltdown continues to cast its shadow not just on current employment and home values but also on longer term issues such as the ability of governments and private organizations to live upto their pension and healthcare liabilities/ obligations. Regardless of whether we choose to acknowledge it or not, the fact remains that a lot of our old assumptions and beliefs about how the world works, including the bedrock currency arrangement coded in the Bretton woods II system, may no longer hold anymore.
Sample these articles -
The end of retirement (from The Economist, June 2010)
Jobs, pensions and markets (from The Economist, Aug 2010)
This has huge implications for us ordinary people as well, even in countries like India which have (so far) been less affected by the crisis.
Unlike Indian government employees, whose retirement and lifetime pension is guaranteed by the Government of India - which because it prints its own money is technically safe from bankruptcy and default - the rest of us cannot really say our retirement and pensions are guaranteed in any real measure. Many young people in the US for instance question why they have to pay into social security when chances are they will never get to see a penny from there by the time they retire. We are indeed in the vanguard of the generations that will have to plan for their futures in exacting detail in the backdrop of increasing uncertainty in the years to come.
That automatically brings up its own host of interesting questions - How will we do? What will do? More generally, what is the current crop of professionals in their 20s and 30s thinking and planning in this scenario? How much confidence do they have in the current set of assets and institutions to protect their savings against inflation? What are their concerns, their preferences and their constraints? How and how much does their view and their actual investment pattern depend on their age, gender, wealth, life cycle stage, nature of employment or other factors?
Mind you, this is not idle chatter - In the aggregate, Indian household savings have averaged about 32-35% of GDP annually in recent years. Unprecedented, really. And thanks to all the visionaries starting with Shri PVN Rao and onto Shri AB Vajpayee and then Shri Manmohan Singh for enabling this. Coupled with an Incremental Capital Output Ratio (ICOR) of about 4 to 1, we are essentially assured growth of about 8% p.a. in real terms for at least the next few years based on sustainable internal demand alone. The choices that our target population segment - the young upwardly mobile urban middle class - will make will be multiplied several times fold as these people tend to be opinion leaders and trend setters also. Where will this mass of money flow to - which asset classes? In what proportion? With what confidence should conditions change down the line? So many questions! The project is one attempt to bring some structure and order to this confusing mass of questions by the application of the scientific method and statistical analysis to understand this conundrum. OK, that is the big picture view of a very complex issue. You've already read the scope document, so you now also have a good handle on the actual problem XYZ corp is handling.
Now, kindly read through this post made here in full if you haven't already:
http://marketing-yogi.blogspot.com/2010/10/some-phase-i-guidance.html
This summarizes my initial views on how to tackle a question as complex as this - basically that:
(i) we limit the scope to what can be done realistically. In this case, we focus only on the big-picture, only at the broad asset class level and not go into details of asset sub-classes and particular schemes available therein. I chose to focus broadly only on Equities (including MFs, SIPs and demat equities), Fixed income plans (bonds, insurance policies that fall under this purview, certain pension and annuity plans etc), property (different types and in different types of cities)
(ii) we focus only on the distribution of asset class preferences in the target population - perhaps as functions of demographics and psychographics.
(iii) we come up with recommendations for which asset classes to go with - some sort of ranking perhaps - given the preference structure of the target population.Any secondary analysis that uses population sizes for such segments (e.g. from the Marketing Whitebook) would be a great plus, IMHO.
OK. This post is at a general, big-picture level and is long enough as-is. Shall continue with more specific guidance in subsequent posts.
Again, not all of you may find general gyan of this sort immediately useful. That is fine. Feel free to go with whatever you have in mind. However, have someone in the group look for updates on the specific guidance.
Sudhir
Thursday, November 4, 2010
Some quick announcements
Hi all,
A few quick announcements.
1. The deadline for phase III is Wednesday (10-11-2010) morning 6 am. A dropbox will be created and opened on blackboard for this purpose. Your deliverable for the project will be a power-point presentation with an upper limit of 30 slides for the main presentation and no limit for auxiliary or backup slides
2. Lecture 10 will essentially be student group presentations of their project report. Kindly ensure the entire group is present in class for this purpose. Groups will be picked up at random. Presenting groups will have 15 minutes of talk time + 3 minutes of Q&A from the class at large. My colleague, prof. Tanuka Ghoshal, who once worked with IMRB Calcutta (real MKTR experience, folks!) has kindly agreed to join the panel evaluating the presentations.
OK, a few groups, maybe 4-5 can be accommodated in class, what about the rest? SO, I'll ask the other groups to take a 20 minute appointment with me and make the presentations. Shall book a conference room/MLT for the purpose. This is to ensure all groups get a fair chance to showcase their PPts, their thinking and emphasize their recommendations to XYZ.
I'll create a google spreadsheet on view access which I will fillup as you email me your preferred slots for. I'm told that since many folks have travel plans for the next weekend, I'd rather get presentations for groups of sections A and B done on Wednesday (10-Nov) itself in the evening, and those for sections C & D on 11-Nov in the evening.
3.Pls find below the lastest update on group respondent collection status. 2951 responses in and still counting, folks! This is exciting. Even dicounting say, 10% to within ISB multiple surveys by a single respondent and another 20% to incomplete/unusable survey, we still end up with a neat 2000 odd usable responses. Wow. Great going there.
4. I've been asked to write up the logic behind the way the final questionnaire has been structured. I shall attempt to provide a write-up for my thinking when I put this together, with ample help from the AAs. A lot of this hinges on what was written in the scope document and the interpretation of which was further expanded upon in this blogpost:
http://marketing-yogi.blogspot.com/2010/10/some-phase-i-guidance.html
Kindly read the above post in full if you haven't already. It pretty much represents the tack I have tried to take in designing the phase II survey. However, you needn't feel constrained with my thought process and could feel free to use the info collected to support the story you may have had in mind a priori. I shall do this over the next few days.
Even otherwise, I plan to dig up and post useful info relating to what last year's class got right and wrong regarding their project in phase III last year. Now, there's no reason why you should repeat mistakes already made last year. Hence, I would certainly recommend that at least a few people from each group keep an eye on the blog for updates in this regard.
5. I was thinking of holding an 'open' tutorial on JMP, methods, anything project related etc over the weekend sometime. Friday is diwali, sat and sun, am told many folks are traveling home for the weekend. Well, I'll try to make this monday evening sometime for roughly an hour then.
6. You may have noticed already that the final slides putup on blackboard may have some slides that either weren't covered in a particular section or have been altered significantly since. This happens because, well, the slides are evolving as the course proceeds. I would recommend that folks download the latest set of slides uploaded at the end of the lecture for section 'D' for use later on. Of course, I try to ensure earlier sections do not lose out on important things that may have come up later on. One reason for having the blog is to be able to communicate interesting/important issues arising in one section across the entire class. LIke happened in the JMP-ANOVA case in sections A and B, sometimes, it would make sense to actually re-do a portion of the slides in the subsequent lecture.
Chalo, last but not least,
"A HaJaAr HaPpY DiWaLi To YoU AnD FaMiLy!"
Sudhir
A few quick announcements.
1. The deadline for phase III is Wednesday (10-11-2010) morning 6 am. A dropbox will be created and opened on blackboard for this purpose. Your deliverable for the project will be a power-point presentation with an upper limit of 30 slides for the main presentation and no limit for auxiliary or backup slides
2. Lecture 10 will essentially be student group presentations of their project report. Kindly ensure the entire group is present in class for this purpose. Groups will be picked up at random. Presenting groups will have 15 minutes of talk time + 3 minutes of Q&A from the class at large. My colleague, prof. Tanuka Ghoshal, who once worked with IMRB Calcutta (real MKTR experience, folks!) has kindly agreed to join the panel evaluating the presentations.
OK, a few groups, maybe 4-5 can be accommodated in class, what about the rest? SO, I'll ask the other groups to take a 20 minute appointment with me and make the presentations. Shall book a conference room/MLT for the purpose. This is to ensure all groups get a fair chance to showcase their PPts, their thinking and emphasize their recommendations to XYZ.
I'll create a google spreadsheet on view access which I will fillup as you email me your preferred slots for. I'm told that since many folks have travel plans for the next weekend, I'd rather get presentations for groups of sections A and B done on Wednesday (10-Nov) itself in the evening, and those for sections C & D on 11-Nov in the evening.
3.Pls find below the lastest update on group respondent collection status. 2951 responses in and still counting, folks! This is exciting. Even dicounting say, 10% to within ISB multiple surveys by a single respondent and another 20% to incomplete/unusable survey, we still end up with a neat 2000 odd usable responses. Wow. Great going there.
4. I've been asked to write up the logic behind the way the final questionnaire has been structured. I shall attempt to provide a write-up for my thinking when I put this together, with ample help from the AAs. A lot of this hinges on what was written in the scope document and the interpretation of which was further expanded upon in this blogpost:
http://marketing-yogi.blogspot.com/2010/10/some-phase-i-guidance.html
Kindly read the above post in full if you haven't already. It pretty much represents the tack I have tried to take in designing the phase II survey. However, you needn't feel constrained with my thought process and could feel free to use the info collected to support the story you may have had in mind a priori. I shall do this over the next few days.
Even otherwise, I plan to dig up and post useful info relating to what last year's class got right and wrong regarding their project in phase III last year. Now, there's no reason why you should repeat mistakes already made last year. Hence, I would certainly recommend that at least a few people from each group keep an eye on the blog for updates in this regard.
5. I was thinking of holding an 'open' tutorial on JMP, methods, anything project related etc over the weekend sometime. Friday is diwali, sat and sun, am told many folks are traveling home for the weekend. Well, I'll try to make this monday evening sometime for roughly an hour then.
6. You may have noticed already that the final slides putup on blackboard may have some slides that either weren't covered in a particular section or have been altered significantly since. This happens because, well, the slides are evolving as the course proceeds. I would recommend that folks download the latest set of slides uploaded at the end of the lecture for section 'D' for use later on. Of course, I try to ensure earlier sections do not lose out on important things that may have come up later on. One reason for having the blog is to be able to communicate interesting/important issues arising in one section across the entire class. LIke happened in the JMP-ANOVA case in sections A and B, sometimes, it would make sense to actually re-do a portion of the slides in the subsequent lecture.
Chalo, last but not least,
"A HaJaAr HaPpY DiWaLi To YoU AnD FaMiLy!"
Sudhir
Do believe the hype
Well, was browsing through this typical Tom Friedman column titled 'Do believe the hype' - IMHO the column is rather shallow high on anecdotes and low on evidence - when I chanced upon this beautiful example of hands-on innovation aiding our rural masses in meaningful ways. I've always been a hard proponent of the mobile revolution changing business, possibilities and ecosystems and this just comes as more sweet vindication (or selective memory, whichever way you see it).
Do Believe the Hype (By THOMAS L. FRIEDMAN in NYT)
Wow. Eh? They broke even. Awesome. May hajaar more such flowers and stories bloom, some led by ISB grads, I hope. So, what was the investment that went in? Where did the idea come from? Who are these young geeks out to change India and thereby the world?
Wow. By now, my admiration for this duo is sky high already. Think of what an incorruptible biometric identification universal ID like the UID can do in this milieu? Thin of the ecosystem for business it can create and spawn.....wow only.
Oh, might as well. On the same note is this splendid piece from Business standard by Arvind Singhal. Recommended read again and a better researched article than Friedman's.
The next five years (in the Indian Economy)
I believe it was some European CXO on a Des visit who remarked "A day in India is like a week in Europe." Yup, I guess.
Wow, eh? Jai ho, indeed.
But, as is often the case with Yindia, the challenges rise just as fast as the opportunities....only.
And so on and on. The next decade is crucial folk.s Make or break is it. We have to, and I mean have to, make the big push and enter the club of solid middle-income countries by 2020. There is no way in hell or highwater that we can afford to miss this bus, as we have done so often in the past.
Jai ho and Jai Bharat.
Do Believe the Hype (By THOMAS L. FRIEDMAN in NYT)
Here’s an example of why I ask these questions. It’s a typical Indian start-up I visited in a garage in South Delhi, EKO India Financial Services. Its founders, Abhishek Sinha and his brother Abhinav, began with a small insight — that low-wage Indian migrant workers flocking to Delhi from poorer states like Bihar had no place to put their savings and no secure way to send money home to their families. India has relatively few bank branches for a country its size, so many migrants stuff money in their mattresses or send cash home through traditional “hawala,” or hand-to-hand networks.
The brothers had an idea. In every Indian neighborhood or village there’s usually a mom-and-pop kiosk that sells drinks, cigarettes, candy and a few groceries. Why not turn each one into a virtual bank? So they created a software program whereby a migrant worker in Delhi using his cellphone, and proof of identity, could open a bank account registered on his cellphone text system. Mom-and-pop shopkeepers would act as the friendly neighborhood local banker and do the same.
Then the worker in New Delhi could give a kiosk owner in his slum 1,000 rupees (about $20), the shopkeeper would record it on his phone and text receipt of the deposit to the system’s mother bank, the State Bank of India. Then the worker’s wife back in Bihar could just go to the mom-and-pop kiosk in her village, also tied into the system, and make a withdrawal using her cellphone. The shopkeeper there would give her the 1,000 rupees sent by her husband. Each shopkeeper would earn a small fee from each transaction. Besides money transfers, workers could also use the system to bank their savings.
Since opening 18 months ago, their virtual bank now has 180,000 users doing more than 7,000 transactions a day through 500 “branches” — mom-and-pop kiosks — in Delhi and 200 more in Bihar and Jharkhand, the hometowns of many maids and migrants. EKO gets a tiny commission from the Bank of India for each transaction and two months ago started to turn a small profit.
Wow. Eh? They broke even. Awesome. May hajaar more such flowers and stories bloom, some led by ISB grads, I hope. So, what was the investment that went in? Where did the idea come from? Who are these young geeks out to change India and thereby the world?
Abhishek, who was inspired by a similar program in Brazil, said the kiosk owners “are already trusted people in each community” and are already in the habit of extending credit to their poor customers: “So we said, ‘Why not leverage them?’ We are the agents of the bank, and these retailers are our subagents.” The cheapest cellphone today has enough computing power to become a digital “mattress” and digital bank for the poor.
The whole system is being run out of a little house and garage with a dozen employees, a bunch of laptops, servers and the Internet. The core idea, says Abhishek, is “to close the last mile — the gap where government services end and the consumer begins.” There is a huge business in bridging that last mile for millions of poor Indians — who, without it, can’t get proper health care, education or insurance.
Wow. By now, my admiration for this duo is sky high already. Think of what an incorruptible biometric identification universal ID like the UID can do in this milieu? Thin of the ecosystem for business it can create and spawn.....wow only.
Aha. Welcome home, Signor. Lest we forget America was made great also by a large swarm of skilled immigrants from all over the world - from German aerospace pioneers to desi silicon valley stars. India, traditionally has been this open land where ideas and immigrants from all over the world could come and mix and create and bedazzle. We lost that mojo somewhere in the past 1000 years but thanks to beginning made on 15 Aug 1947, we're well on our way to regain our traditional influence and place in the world, Heaven willing of course.
What is striking about the small EKO team is that it includes graduates from India’s most prestigious institutes of technology who were working in America but decided to come home for the action, while the chief operating officer — Matteo Chiampo — is an Italian technologist who left a good job in Boston to work here “where the excitement is,” he said.
India today is this unusual combination of a country with millions of people making $2 and $3 a day, but with a growing economy, an increasing amount of cheap connectivity and a rising number of skilled technologists looking to make their fortune by inventing low-cost solutions to every problem you can imagine. In the next decade, I predict, we will see some really disruptive business models coming out of here — to a neighborhood near you. If you thought the rate of change was fast thanks to the garage innovators of Silicon Valley, wait until the garages of Delhi, Mumbai and Bangalore get fully up to speed. I sure hope we’re ready.OK. This is classic Friedman. Feel-good and all but better not to get carried away prematurely only. Actions speak louder than words and let the deeds of our entrepreneurs speak in making our country a better place for all its citizens to live in, while doing better than breaking even themselves.
Oh, might as well. On the same note is this splendid piece from Business standard by Arvind Singhal. Recommended read again and a better researched article than Friedman's.
The next five years (in the Indian Economy)
I believe it was some European CXO on a Des visit who remarked "A day in India is like a week in Europe." Yup, I guess.
India’s GDP in 2005 was about $785 billion, which has increased to almost $1,400 billion in 2010 and should cross $2,000 billion by 2015, implying that Indian economy would have added almost one India of 2005 to its size in the next five years. This spectacular growth, on a relatively high base of almost $1.4 trillion, translates into unprecedented opportunities and greater challenges than what India has faced in the past. Let us look at 2005 and then try to crystal ball 2015 for a few consumption categories.
Wow, eh? Jai ho, indeed.
In 2005, our domestic market for four-wheelers was just about 1.1 million vehicles. It will cross 2 million in 2010 and touch 4 million by 2015 (including as many as 110,000 in the luxury and premium segments). There were less than 100,000 users of smart-phones in 2005, about 25 million in 2010, and perhaps as many as 225 million in 2015. From a user base of about 8 million laptop users in 2010, there could be as many as 50 million by 2015. There were less than 25 million Internet users in 2005, increasing to about 75 million in 2010, and projected to grow to more than 250 million by 2015. From less than 250 multiplex screens in 2005, to about 800 in 2010, the count could cross 2,000 by 2015. Similar increases in multiples ranging from 2x-10x could happen across many consumer product categories.
..... Since the market opportunity is likely to grow in multiples of the current size, most businesses should rework their strategies and rewrite their business plans rather than just continue with incremental additions to their current plans.
But, as is often the case with Yindia, the challenges rise just as fast as the opportunities....only.
It is, however, more important to also give a hard look at the challenges that Indians and Indian businesses will encounter in the coming years. The first and more critical is the physical infrastructure. Doubling of number of cars and two-wheelers on the roads will need a corresponding increase not only in the road network but also in parking spaces, auto repair workshops and fuel vending stations. Adding of almost $400 billion in additional private consumption by 2015 will require a mind-boggling increment of nearly 1.5 billion square feet of retail space alone (or about 6,000 shopping malls of 250,000 square feet each). E-waste from the installed base of hundreds of millions of electrical and digital devices will run into tens of millions of tonnes, requiring enhancement of waste-management capacity by almost 10 times of what exists today. Steady increase in urbanisation will imply an additional demand for almost 20 million urban dwellings in cities where the infrastructure is already creaking to its limits. Strain on social infrastructure, be it education (primary, secondary, higher, and vocational) or health care (primary, secondary, and tertiary) or clean drinking water or sanitation will further increase manifold, rather just in simple double-digit percentages.
And so on and on. The next decade is crucial folk.s Make or break is it. We have to, and I mean have to, make the big push and enter the club of solid middle-income countries by 2020. There is no way in hell or highwater that we can afford to miss this bus, as we have done so often in the past.
Jai ho and Jai Bharat.
Wednesday, November 3, 2010
Websurvey latest news
Hi All,
As of now, some 2243 responses have flowed in. I thin its been great going so far.
Below is a pivot table with the # responses by group I have got at the time of writing. There's some glitch that is disallowing me currently from collecting more than 2000 responses, IT says it can be fixed.
Some concerns:
1. I'm told some ISB folk have filled up multiple surveys on behalf of multiple teams. This was precisely what I feared when allowing ISB folk to fill in surveys.
People, respondents' IP addresses are logged by qualtrics. Multiple surveys by within-ISB IPs will not be counted, pls be aware.
Also, within-ISB surveys are permitted only for foreign or exchange students only. Upto 10 responses from within ISB from each of such students is enough. Additional responses from within ISB for groups may not be counted for phase II, please be cautioned.
2. The fact that the final survey was hurried through and had to be released sans requisite pre-testing shows in that there are several ways in which the questionnaire could be improved. That said, however, we go with what we currently have. Let us make the best of what is out there currently.
Shall add more thoughts as they occur.
Sudhir
As of now, some 2243 responses have flowed in. I thin its been great going so far.
Below is a pivot table with the # responses by group I have got at the time of writing. There's some glitch that is disallowing me currently from collecting more than 2000 responses, IT says it can be fixed.
Some concerns:
1. I'm told some ISB folk have filled up multiple surveys on behalf of multiple teams. This was precisely what I feared when allowing ISB folk to fill in surveys.
People, respondents' IP addresses are logged by qualtrics. Multiple surveys by within-ISB IPs will not be counted, pls be aware.
Also, within-ISB surveys are permitted only for foreign or exchange students only. Upto 10 responses from within ISB from each of such students is enough. Additional responses from within ISB for groups may not be counted for phase II, please be cautioned.
2. The fact that the final survey was hurried through and had to be released sans requisite pre-testing shows in that there are several ways in which the questionnaire could be improved. That said, however, we go with what we currently have. Let us make the best of what is out there currently.
Shall add more thoughts as they occur.
Sudhir
Tuesday, November 2, 2010
Websurvey Updates
Hi all,
The websurvey has already garnered some 1100+ responses.
Yay and Whew!
Great effort so far, teams. Let me hurry to add however that the above 1100 responses include quite a few by groups taking the surveys themselves to check the websurvey out.
Thanks to all those who wrote in with pointers to typos and other errors. We've accommodated what changes we could without changing the structure of the survey and invalidating past responses to particular questions.
How are groups doing? Well, here's a picture from a quick pivot I drew as of 1740 hrs Tuesday. Let me see if this comes up well....
Alright. There you go.
Update: Have been asked for a word doc copy of the survey. The argument is that this way groups can study the questionnaire logic and better plan the analysis to come. I agree. I'll do one better, let me share with you the exec summary sheet that has been generated also. Shall ask the AAs to putup on the blackboard.
Sudhir
The websurvey has already garnered some 1100+ responses.
Yay and Whew!
Great effort so far, teams. Let me hurry to add however that the above 1100 responses include quite a few by groups taking the surveys themselves to check the websurvey out.
Thanks to all those who wrote in with pointers to typos and other errors. We've accommodated what changes we could without changing the structure of the survey and invalidating past responses to particular questions.
How are groups doing? Well, here's a picture from a quick pivot I drew as of 1740 hrs Tuesday. Let me see if this comes up well....
Alright. There you go.
Update: Have been asked for a word doc copy of the survey. The argument is that this way groups can study the questionnaire logic and better plan the analysis to come. I agree. I'll do one better, let me share with you the exec summary sheet that has been generated also. Shall ask the AAs to putup on the blackboard.
Sudhir
Monday, November 1, 2010
Websurvey Live Link
Hi Class,
Please find below the websurvey live link.
Pls forward to as many genuine respondents in the target population as you can find. KIndly do not forget to mention two things:-
1. Your group name. The respondent has to enter the name of the group that has forwarded the survey to him/her.
2. The incentives for survey completion - both monetary and data/information related.
If you detect any obvious issues with anything in the survey - typos, logical errors etc, let us know. There is cope to correct typos etc even while the survey is open.
Pls remember, the survey closes midnight Thursday (11.59 pm on 4-Nov-10). All valid and complete responses from each group obtained till then are counted towards the group's credit for phase II.
Sudhir
Errata from Lecture 7
Hi Class,
Lecture 7 deals with Causal research (Experimentation) and the analysis of variance. The two are together because they are connected - the former collects data which the latter is very well suited to analyze for inference, recommendations and action.
The hands-on work regarding running ANOVA models on different software - primarily JMP, Excel and R - in class today led to two errors for which reason I write this erratum online.
1. JMP by default seems to read all data in as numeric only. In ANOVA, our independent variables, the Xs, are nominal/ categorical. So, JMP's reading them wrong resulted in the results being misleading as well. Even the ANOVA results JMP displays are nothing but the joint significance F-test which is computed on a linear model considering the Xs as metric.
The way out is to double-click on each column header and redefine the variable type to categorical/nominal. Then the thing runs fine and the results seem plausible.
2. Excel is not well-suited to doing ANOVA related work primarily because the data input it requires seems to need lots of prior editing.The data must be sorted and placed in arrays before Excel will read and analyze them. JMP is much better placed in this regard. R has its own issues. It does not give the joint significance F-stat and p-value.
So, bottom line - we stick with JMP as the best of the lot - after inputting data in the appropriate manner.
OK, more generally, there are two kinds of statistical tests that we do to conclude things one way or the other - tests of association that typically involve interdependent data, and those of dependence that involve dependent data. The stats primer started with basic model building - waded into over-identified systems and from there into models of dependence. The chi-square tests we saw last class were tests of association and today, we see a powerful battery of tests for dependence.
This has applications on the ground - not just in the project but for business problems, situations and issues you are likely to face at work. Take a re-look at your ELPs to see if the methods picked up could have come in handy. The reason I'm hard-selling data work and quant methods and what we have been doing in the past 2 odd lectures is that I get the vibe I haven't convincingly conveyed their importance, practical utility and breadth of applicability in real-life problems to a significant section of the class.
Well, class, the going gets tougher, of sorts, here on. We're entering a rather dry part of the course involving analysis and some data work. My experience tells me that data work imparts real skills that are sale-able in the marketplace - whether explicitly or implicitly. It lends realism, credibility and perspective to the problem formulation and research design skills we learned in the early part of the course. So even if the lecture appears rather dry, dull and all, hang on. It will all connect together, not least through the project, to tangible knowledge learned and skills earned in the course of the course.
My apologies to sections A and B for the JMP issues that came up. My thanks to the folks who spoke up and pointed out the errors.
Sudhir
Lecture 7 deals with Causal research (Experimentation) and the analysis of variance. The two are together because they are connected - the former collects data which the latter is very well suited to analyze for inference, recommendations and action.
The hands-on work regarding running ANOVA models on different software - primarily JMP, Excel and R - in class today led to two errors for which reason I write this erratum online.
1. JMP by default seems to read all data in as numeric only. In ANOVA, our independent variables, the Xs, are nominal/ categorical. So, JMP's reading them wrong resulted in the results being misleading as well. Even the ANOVA results JMP displays are nothing but the joint significance F-test which is computed on a linear model considering the Xs as metric.
The way out is to double-click on each column header and redefine the variable type to categorical/nominal. Then the thing runs fine and the results seem plausible.
2. Excel is not well-suited to doing ANOVA related work primarily because the data input it requires seems to need lots of prior editing.The data must be sorted and placed in arrays before Excel will read and analyze them. JMP is much better placed in this regard. R has its own issues. It does not give the joint significance F-stat and p-value.
So, bottom line - we stick with JMP as the best of the lot - after inputting data in the appropriate manner.
OK, more generally, there are two kinds of statistical tests that we do to conclude things one way or the other - tests of association that typically involve interdependent data, and those of dependence that involve dependent data. The stats primer started with basic model building - waded into over-identified systems and from there into models of dependence. The chi-square tests we saw last class were tests of association and today, we see a powerful battery of tests for dependence.
This has applications on the ground - not just in the project but for business problems, situations and issues you are likely to face at work. Take a re-look at your ELPs to see if the methods picked up could have come in handy. The reason I'm hard-selling data work and quant methods and what we have been doing in the past 2 odd lectures is that I get the vibe I haven't convincingly conveyed their importance, practical utility and breadth of applicability in real-life problems to a significant section of the class.
Well, class, the going gets tougher, of sorts, here on. We're entering a rather dry part of the course involving analysis and some data work. My experience tells me that data work imparts real skills that are sale-able in the marketplace - whether explicitly or implicitly. It lends realism, credibility and perspective to the problem formulation and research design skills we learned in the early part of the course. So even if the lecture appears rather dry, dull and all, hang on. It will all connect together, not least through the project, to tangible knowledge learned and skills earned in the course of the course.
My apologies to sections A and B for the JMP issues that came up. My thanks to the folks who spoke up and pointed out the errors.
Sudhir
Sunday, October 31, 2010
Phase II questionnaire travails
Wow.
The project is even more challenging / tough than I'd first anticipated. First off, 'good show' to a lot of teams that came up with credible websurveys of their own.
The job of patching together a 'common denominator' questionnaire for phase II by picking up the 'best' bits from everywhere isn't easy. Lots of culling and cutting out and rewriting and editing and everything all over the place. And at the end of all that, am not sure what the final product will look like. Time will tell.
And the fact that we lost precious time in simple admin things like group-formation are now coming home to roost. This sort of thing would ideally require more time and thought.
I hope to be done early and on time. Would like to have phase II launch by as early as Tuesday, hopefully. Let us see.
At present, the survey has 34 questions from 8 sections (including demographics). Have attempted to look more at 'savings in the long-term including retirement planning' rather than retirement planning per se as some may confuse one as very different from the other and get the questions wrong. Many products are being sold through banks as a conduit, it seems. The options available are so wide and colorful, its amazing and disconcerting as well. Am focussing on 3-4 major asset classes - Equities including MFs and SIPs, Housing, Insurance and fixed income annuities including pension plans, FDs and Gold.
Shall post more updates about my thoughts as progress happens.
Sudhir
The project is even more challenging / tough than I'd first anticipated. First off, 'good show' to a lot of teams that came up with credible websurveys of their own.
The job of patching together a 'common denominator' questionnaire for phase II by picking up the 'best' bits from everywhere isn't easy. Lots of culling and cutting out and rewriting and editing and everything all over the place. And at the end of all that, am not sure what the final product will look like. Time will tell.
And the fact that we lost precious time in simple admin things like group-formation are now coming home to roost. This sort of thing would ideally require more time and thought.
I hope to be done early and on time. Would like to have phase II launch by as early as Tuesday, hopefully. Let us see.
At present, the survey has 34 questions from 8 sections (including demographics). Have attempted to look more at 'savings in the long-term including retirement planning' rather than retirement planning per se as some may confuse one as very different from the other and get the questions wrong. Many products are being sold through banks as a conduit, it seems. The options available are so wide and colorful, its amazing and disconcerting as well. Am focussing on 3-4 major asset classes - Equities including MFs and SIPs, Housing, Insurance and fixed income annuities including pension plans, FDs and Gold.
Shall post more updates about my thoughts as progress happens.
Sudhir
Friday, October 29, 2010
Good luck.
Update: Alrite, Pls feel free to use this post's comment section for post-midterm post-mortems, if any.
-------------------------------------
[Good luck] for the midterms.
Well, by the time this post goes up, you'd all be in the exam hall only.
Good luck and Godspeed.
Sudhir
-------------------------------------
[Good luck] for the midterms.
Well, by the time this post goes up, you'd all be in the exam hall only.
Good luck and Godspeed.
Sudhir
Wednesday, October 27, 2010
Tutorial Thu 9 pm AC2
Update: Some folks have class right till 9.15 pm tomorrow, am told. Well, at most I can move the tutorial to 9 pm. I suspect by the time things get warmed up it will already be past 9.15 pm. So, 9 pm tomorrow it ia, folks.
Hi All,
Tutorial happening. Pivots and SOLVER will be covered. Any other midterm queries are also fine. Bring your laptops. Might upload some more datasets for practice.
Sudhir
P.S.
Update:
Pls enter the URL of your's group's survey in the google form given at this link:
http://marketing-yogi.blogspot.com/2010/10/phase-i-submissions.html
Hi All,
Tutorial happening. Pivots and SOLVER will be covered. Any other midterm queries are also fine. Bring your laptops. Might upload some more datasets for practice.
Sudhir
P.S.
Update:
Pls enter the URL of your's group's survey in the google form given at this link:
http://marketing-yogi.blogspot.com/2010/10/phase-i-submissions.html
Midterm related Clarifications
Hi all,
Some quick clarifications re the midterms.
1. There are 3 parts to the midterm - an MCQ part, a short answer part and a laptop based part with marks 25, 35 and 40 respectively. They will also get 30, 45 and 45 minutes respectively. Only the last part is open-book, open-notes whereas the first 2 parts are open-coursepack and cheat-sheet only.
2. Nothing not covered or mentioned in the slides will come for the exam. Look at the textbook only if some slide covered concept is unclear for any reason.The textbook is more for reference - you should know what to look for and where when faced with MKTR problems down the line at work.
3. Hypothesis testing as covered in Lecture 6 will not be covered in the midterm. SOLVER and PIVOT however are very much within bounds.
Any further midterm Q&A, kindly use the comments section below.
Some quick clarifications re the midterms.
1. There are 3 parts to the midterm - an MCQ part, a short answer part and a laptop based part with marks 25, 35 and 40 respectively. They will also get 30, 45 and 45 minutes respectively. Only the last part is open-book, open-notes whereas the first 2 parts are open-coursepack and cheat-sheet only.
2. Nothing not covered or mentioned in the slides will come for the exam. Look at the textbook only if some slide covered concept is unclear for any reason.The textbook is more for reference - you should know what to look for and where when faced with MKTR problems down the line at work.
3. Hypothesis testing as covered in Lecture 6 will not be covered in the midterm. SOLVER and PIVOT however are very much within bounds.
Any further midterm Q&A, kindly use the comments section below.
Tuesday, October 26, 2010
Ice-cream dataset homework related thread
Hi All,
I received the following email from Ipsita:
I haven't solved the homework myself, so shall appreciate your inputs and crowd wisdom in solving the same. Am still working on the midterm paper here.
Sudhir
I received the following email from Ipsita:
If you agree/disagree or better still, have your own answers to share, kindly do so in the comments section of this thread.
Dear Professor,
I have some doubts regarding the answers put on ice cream survey. Although most of the answers match, the following do not.
· Ques 1: iii and iv· Ques 3: Edy's is the biggest competitor to Wows, however the uploaded answer shows Turkey Hill.· Ques 4: ii (173 fall in the $100,000+ household income, i.e. 34.81%)
Could you please clarify on the same.
I haven't solved the homework myself, so shall appreciate your inputs and crowd wisdom in solving the same. Am still working on the midterm paper here.
Sudhir
General announcements & miscell stuff
Hi all,
A few quick announcements:
1. So it turns out that even the esurveyspro.com free account no longer honors SKIP logic (it did perfectly last year) AND that our LRC has discontinued their surveymoneky.com professional subscription. The school has since migrated to another websurvey package called qualtrics. I've spoken with IT and they say you can make your own account with your ISB email ID. Kindly do so and program your survey there - I suspect it should be pretty much akin to surveymonkey.com/ esurveyspro.com only.
This also means that the *final* questionnaire I'll make will be in qualtrics only, and phase II will be administered in qualtrics.
This is the email IT sent me:
2. Was asked in class how to subscribe to blog feed for post alerts etc. Vignesh of sec 'D' has been kind enough to share one procedure with the class. I'm posting his email in full.
3. Kindly ensure JMP is up and running in your laptops for lecture 6. Might as well try to refresh JMP usage while we are at work on the car survey raw data. Pls also ensure Excel analysis toolpak (that contains Excel's regression function) is up and loaded in your machines.
4. Well, came across this splendid article the other day. IMO, its worthwhile sharing with the class.
On Trading and Markets
I'd say the entire article, based on a speech made last week Thursday in NYC is a good read, but to save time, let me excerpt the course relevant bits here...
5. Was on the panel for an ITC case contest yesterday. In my interactions with the ITC brass, I told them I teach MKTR and I asked for their thoughts on the current output at B-schools, they said and let me quote as much verbatim as I can:
6. There are folks who come up with good, relevant questions etc after class or in the break and when I ask them to post the same as comments on the blog so that I can reply and the entire class can share, they miraculously disappear. IMHO, positive contributions on the blog should merit some credit - if only as incentive to be a little more open and share thoughts with the entire class.
Sudhir
A few quick announcements:
1. So it turns out that even the esurveyspro.com free account no longer honors SKIP logic (it did perfectly last year) AND that our LRC has discontinued their surveymoneky.com professional subscription. The school has since migrated to another websurvey package called qualtrics. I've spoken with IT and they say you can make your own account with your ISB email ID. Kindly do so and program your survey there - I suspect it should be pretty much akin to surveymonkey.com/ esurveyspro.com only.
This also means that the *final* questionnaire I'll make will be in qualtrics only, and phase II will be administered in qualtrics.
This is the email IT sent me:
Hope that helps. Once you have the word doc questionnaire ready, programming into qualtrics shouldn't be too much time or trouble, hopefully.
Dear Professor,
Please find the link below for registering to Qualtrics survey.
Students can register to the same using their ISB email id.
isb.qualtrics.com
please let me know for any further help
Regards
Vijaya Lakshmi
Executive - IT Applications
2. Was asked in class how to subscribe to blog feed for post alerts etc. Vignesh of sec 'D' has been kind enough to share one procedure with the class. I'm posting his email in full.
For both Outlook 2007 and 2010, the process is fairly simple –
1. Copy the RSS feed link from the blog – in the case of Marketing Yogi, the RSS feed is at http://marketing-yogi.blogspot.com/feeds/posts/default. Please note that this is a feed for only the posts on the blog.
2. In Outlook 2007/2010, right click on RSS Feeds in the left navigation bar, select Add a New RSS Feed and enter the feed address from above.
3. Click Advanced and select the option to download enclosures (any images in the feed for example) as well as the option to download the full article as an .html attachment.
4. Click OK, then Yes on the previous window and you’re done.
Thanks,
Best,
Vignesh.
3. Kindly ensure JMP is up and running in your laptops for lecture 6. Might as well try to refresh JMP usage while we are at work on the car survey raw data. Pls also ensure Excel analysis toolpak (that contains Excel's regression function) is up and loaded in your machines.
4. Well, came across this splendid article the other day. IMO, its worthwhile sharing with the class.
On Trading and Markets
I'd say the entire article, based on a speech made last week Thursday in NYC is a good read, but to save time, let me excerpt the course relevant bits here...
Good evening, ladies and gentlemen. My name is Ben Davies ...
Well, nice to see the 'unknown-unknown' find its origin, form and function out there, eh?
Economics has sought to blend epistemology, physics, mathematics, and behavioral science to try to measure uncertainty. They aim to try to predict when we might have an economic collapse, but no model has been created that manages this with much confidence, if any at all. How do you measure a risk that is unmeasurable?
No, there is nothing certain about economic predictions. Donald Rumsfeld, the former U.S. defense secretary, unwittingly declared it so at a NATO press conference in 2002, when he responded to a question on intelligence gathering:
"It's not the certainties that make life interesting; it's the uncertainties. There are known knowns. These are things we know that we know. There are known unknowns. That is to say, there are things we know we don't know. But there are also unknown unknowns -- the things we don't know we don't know."
"Unknown unknowns" -- at the time this was ridiculed as a piece of deliberate and meaningless obfuscation. Rumsfeld even won an award from the British Plain English Campaign for the most nonsensical remark made by a public figure. I would add that he narrowly pipped California Gov. Arnold Schwarzenegger, who commented, "I think that gay marriage is something that should be between a man and a woman."
Ridicule aside, I do think it was brilliant piece of polemic, irrespective of one's political persuasion. Without knowing it, Rumsfeld could actually be the poster child for a new line of economic thought that tries to draw parallels from physics -- the Heisenberg Uncertainty Principle.
In quantum physics this principle states that certain pairs of physical properties, such as position and momentum, cannot be simultaneously known to arbitrarily high accuracy.
And herein lies the issue with macroeconomics. One cannot be deterministic. Events are not causally determined by previous events alone. A known event does not predict a known outcome.
Although much can be learned from this theory, this is where the analogy ends. The act of experimenting with the electron and measuring position did not alter the outcome. In a human system the very act of forecasting or predicting an outcome actually influences that very outcome. This we call the feedback mechanism. It is encompassed in a theory many will know as "reflexivity."
5. Was on the panel for an ITC case contest yesterday. In my interactions with the ITC brass, I told them I teach MKTR and I asked for their thoughts on the current output at B-schools, they said and let me quote as much verbatim as I can:
"You chaps train people to be Marketing vice-presidents, not territory sales managers or brand managers. Many recent MBAs don't know the basic of the craft. How to evaluate the effectiveness of an ad campaign, commission what type of market research with which agency or analyze a Nielsen sales report. You teach the concept but not the craft of marketing unlike engineers or chartered accountants who learn the craft first."Or some such flow. Well, I told them my students can do all that and more (or should be able to by the end of lec 10). And yes, it did feel like sweet vindication, external validation if you will - we're doing the practical stuff, the useful stuff. Good. The other day, one Mr, Nithin Vyakarnam, an entrepreneur alum from '06, who sits in on the class with section C also mentioned similar things - that very few MBA folks out in the real world know how to handle pivots or run models. So, we are doing something good after all. Hopefully. Time will tell, I guess.
6. There are folks who come up with good, relevant questions etc after class or in the break and when I ask them to post the same as comments on the blog so that I can reply and the entire class can share, they miraculously disappear. IMHO, positive contributions on the blog should merit some credit - if only as incentive to be a little more open and share thoughts with the entire class.
Sudhir
Monday, October 25, 2010
Caselet 5.1.
Hi All,
We had caselet 5.1. read and discussed in class today (sections A & B). I get the feeling it didn't go as well as I'd hoped. Essentially, the points regarding the caselet I'd wanted to articulate likely did not get across to the class.
Let me try to make amends and write down the main issues in that caselet here, separately.
1. Tech innovation and adoption are often (but not always) a function of the installed base for an application in any market. Hence, breakthroughs in broadband use and businesses built around broadband applications are most likely to come from South Korea than from anywhere else. One big point Shri Baijal makes in the article is that India has already acquired such a technological installed base in telecom and the next big innovations may get implemented here as a result. That is no small shakes.
2. Given that the desi telecom mart may become the testbed of choice for new technologies and business models in the telecom space, the author then goes onto speculate about the effects of what 4G or 5G means in speed terms. The impact, I thought, was simply wow. The amount of potential for disruption, innovation and basic change in the lives of ordinary people I could see there simply blew me over. However, I doubt my enthusiasm carried over as articulation.
3. OK, what was caselet 5.1. doing in the qualitative research section? Recall caselet 2.2 where we first discussed cross-industry disruptive influences. This is similar. From music to movies to education to medicine to personal communication to networking - every big sector I could think of from a consumer standpoint was being impacted. The potential for creation of new ecosystems that supported myriad new businesses and biz models overnight could be seen - the Oasis analogy, I wanted to convey. Then the UID example as an extension of this same phenomenon. End of the day, all this would manifest as latent needs that consumers can't articulate anything about today, requiring methods from the qualitative research toolkit - Indirect questioning, projective techniques, unstructured questioning etc would need to be brought to bear.
Thanks and regards.
Sudhir
We had caselet 5.1. read and discussed in class today (sections A & B). I get the feeling it didn't go as well as I'd hoped. Essentially, the points regarding the caselet I'd wanted to articulate likely did not get across to the class.
Let me try to make amends and write down the main issues in that caselet here, separately.
1. Tech innovation and adoption are often (but not always) a function of the installed base for an application in any market. Hence, breakthroughs in broadband use and businesses built around broadband applications are most likely to come from South Korea than from anywhere else. One big point Shri Baijal makes in the article is that India has already acquired such a technological installed base in telecom and the next big innovations may get implemented here as a result. That is no small shakes.
2. Given that the desi telecom mart may become the testbed of choice for new technologies and business models in the telecom space, the author then goes onto speculate about the effects of what 4G or 5G means in speed terms. The impact, I thought, was simply wow. The amount of potential for disruption, innovation and basic change in the lives of ordinary people I could see there simply blew me over. However, I doubt my enthusiasm carried over as articulation.
3. OK, what was caselet 5.1. doing in the qualitative research section? Recall caselet 2.2 where we first discussed cross-industry disruptive influences. This is similar. From music to movies to education to medicine to personal communication to networking - every big sector I could think of from a consumer standpoint was being impacted. The potential for creation of new ecosystems that supported myriad new businesses and biz models overnight could be seen - the Oasis analogy, I wanted to convey. Then the UID example as an extension of this same phenomenon. End of the day, all this would manifest as latent needs that consumers can't articulate anything about today, requiring methods from the qualitative research toolkit - Indirect questioning, projective techniques, unstructured questioning etc would need to be brought to bear.
Thanks and regards.
Sudhir
Quiz 2 issues
Hi all,
A quiz was planned, based on the Ice-cream dataset for sections A & B today which for various reasons - incluidng ambiguity in the quiz paper and non-correspondence of question numbers in the dataset with Q numbers in the quiz paper - could not be conducted smoothly.
Well, my apologies. We'll have to scratch this quiz now.
But in general, I should make clear that the raw data from the survey has a different question numbering than in the questionnaire because each column corresponds to an answer option. My mistake, I overlooked that fact. Still, using question headers in the top row of the dataset, one could still have continued but then 10 minutes wouldn't have been enough to manage all this in.
I've released last year's 2009 car survey questionnaire and raw data on blackboard. Kindly go through the same. Familiarize yourself with it. Chances are high that questions based on it might landup in the midterm. The questions would be roughly in the same mould as those for the Ice cream and beer datasets, though.
P.S.
We're now entering a more quantitative phase in the course. It gets difficult for me to judge how much already know -> I'm going too slow and where I'm going too fast. Timely, constructive feedback would be nice, I guess.
A quiz was planned, based on the Ice-cream dataset for sections A & B today which for various reasons - incluidng ambiguity in the quiz paper and non-correspondence of question numbers in the dataset with Q numbers in the quiz paper - could not be conducted smoothly.
Well, my apologies. We'll have to scratch this quiz now.
But in general, I should make clear that the raw data from the survey has a different question numbering than in the questionnaire because each column corresponds to an answer option. My mistake, I overlooked that fact. Still, using question headers in the top row of the dataset, one could still have continued but then 10 minutes wouldn't have been enough to manage all this in.
I've released last year's 2009 car survey questionnaire and raw data on blackboard. Kindly go through the same. Familiarize yourself with it. Chances are high that questions based on it might landup in the midterm. The questions would be roughly in the same mould as those for the Ice cream and beer datasets, though.
P.S.
We're now entering a more quantitative phase in the course. It gets difficult for me to judge how much already know -> I'm going too slow and where I'm going too fast. Timely, constructive feedback would be nice, I guess.
Sunday, October 24, 2010
A few quick announcements
Update: FYI and just to clarify:
(i) Use the free account of www.esurveyspro.com for your phase I submission, *not* that of surveymonkey.com. (ii) Do *not* attempt to administer your phase I survey to anyone outside of pre-testing. I'll put together a consolidated, final questionnaire for the entire class based on your phase I submissions, on LRC's survey monkey pro account, that we will use for phase II - data collection. (iii) Pls feel free to ask questions and clarifications. Recall that MKTR involves a collaborative process with client input - I'll play the role of the client in case you need clarification on the scope and objectives etc.
Sudhir
Hi all,
Some quick announcements:
1. If you haven't already, pls take last year's project survey ASAP. Every MKTR student should take it regardless of whether or not his/her group-mates have done so already. I plan to release the questionnaire and the results (summary + raw data) over blackboard soon. I strongly recommend having familiarity with this dataset as well.
*Basically, I cannot guarantee that questions pertaining to this dataset etc will not land up in the midterm.
2. Pls refresh your Excel SOLVER skills. Am told you have done quite a bit of this on DMOP or something. Good, as it saves me a ton of trouble. We're starting out with the Modeling and Solution primer tomorrow for sections A & B. DOn;t forget to bring your laptops, janta.
3. We're racing to complete grading Quiz I as we speak. Would like to clarify re Quiz I for sections C & D that only people who score 0 out of 8 in the first 2 questions will be eligible and be invited to take a re-quiz for those 8 points. If the eligible folks don;t want to the the quiz, fine. This is a one-time exception only. I odn;t want folks to think such will happen after every quiz or anything.
Thanks and regards,
Sudhir
(i) Use the free account of www.esurveyspro.com for your phase I submission, *not* that of surveymonkey.com. (ii) Do *not* attempt to administer your phase I survey to anyone outside of pre-testing. I'll put together a consolidated, final questionnaire for the entire class based on your phase I submissions, on LRC's survey monkey pro account, that we will use for phase II - data collection. (iii) Pls feel free to ask questions and clarifications. Recall that MKTR involves a collaborative process with client input - I'll play the role of the client in case you need clarification on the scope and objectives etc.
Sudhir
Hi all,
Some quick announcements:
1. If you haven't already, pls take last year's project survey ASAP. Every MKTR student should take it regardless of whether or not his/her group-mates have done so already. I plan to release the questionnaire and the results (summary + raw data) over blackboard soon. I strongly recommend having familiarity with this dataset as well.
*Basically, I cannot guarantee that questions pertaining to this dataset etc will not land up in the midterm.
2. Pls refresh your Excel SOLVER skills. Am told you have done quite a bit of this on DMOP or something. Good, as it saves me a ton of trouble. We're starting out with the Modeling and Solution primer tomorrow for sections A & B. DOn;t forget to bring your laptops, janta.
3. We're racing to complete grading Quiz I as we speak. Would like to clarify re Quiz I for sections C & D that only people who score 0 out of 8 in the first 2 questions will be eligible and be invited to take a re-quiz for those 8 points. If the eligible folks don;t want to the the quiz, fine. This is a one-time exception only. I odn;t want folks to think such will happen after every quiz or anything.
Thanks and regards,
Sudhir
Some Phase I Guidance
Update: Got this comment which I replied to in an older post but IMHO it bears wider dissemination.
Rajesh writes:
Warning: Longish post - but worthwhile read, IMHO.
Hi All,
There've been some queries that might be more widely shared by the class at large. Let me use this post to try to answer some.
Vikram writes:
In general, I've avoided giving much 'guidance' on the project for fear of biasing your natural thought processes and responses to the scope document. So let me stress, my ideas aren't necessarily the only 'right' approach to this problem. If your approach differs, fine.
People, please lookup the comments sections to different posts for relevant Q&A. Pls use the comments section for follow-up Q&A. Let me first quote what I wrote in the comments section of 1 post below:
Some folks may not even have thought about retirements and longterm savings. What is the % of such people in the target population? What markers - demographic or psychographic - may distinguish them from other folks?
Then again, some other folks may have thought deeply about this issue and planned their retirements in some detail. What is the % of such folks, their motivations, background, Identification markers, and importantly, their current portfolio like?
Some people would be in-between the two extremes above. How amenable are they to the message that retirement planning is important? How best to reach them via media exposure?
In general, how do people view investments such as property, gold, stocks, etc? Long term savings-vehicles? Inflation hedges? Retirement income sources? I don't know what else you may come up with. These things I am thinking up as I type.
Even more generally, what factors affect the awareness, attitudes, ideas and actions of people vis-a-vis the issue at hand? Is it their age, their stage in the life-cycle, their family size - # children, their employment type, peers and social circle, the influence their parents' choices in this regard had on them, opportunities at hand, etc etc.
If you want to use standard Marketing frameworks you have learnt elsewhere here, be my guest. For instance, I can think of the AIDA framework in Kotler that could be gainfully applied. AIDA stands for 'Awareness-Intention-Decision-Action' and measures where potential consumers are on this continuum vis-a-vis a particular product.Erratum: 'Decision' was previously incorrectly IDed as 'desire'
All these things may or maynot figure in your design of the questionnaire. All of them probably cannot be accommodated anyway in the spacetime constraints we are operating under. This then, is the challenge for you, the call you have to make.
The 15+ person-hours (at least, I expect) your group spends on the questionnaire design should be great leaning in the process of satisficing competing demands and requirements, optimizing available resources within set constraints and all in the framework of a complex issue at hand. This is the real learning the course hopes to impart - the bookish gyan is always there anyway.
Chalo. The post is long enough already. I should stop. Pls feel free to use the comments section for follow-up Q&A, feedback, critique and discussion
Sudhir
P.S. Also meant to mention some best practices from last year that came into handy. Some groups modularized the questionnaire and had subgroups of 2-3 work on separate modules. Apart from the efficiency of parallel-processing that this mechanism gave them, an added advantage was that one subgroup could pre-test the others' modules. Am not sure if this will work for your particular group or not, but something to consider.
Also, let me once again, strongly urge groups to setup and use free Wikis to coordinate activities across time and distance. Pls see the relevant blogpost below for the link to Wiki creation.
Rajesh writes:
Dear ProfessorI respond:
The explanation of the situation in which the management (XYZ corp)currently is appears very short in the project scope document.
Is the XYZ corp trying to figure out whether to enter into various segment like banking/finance/Insurance which will be available to it post the new license policy or these sectors are just some of the options among others like real estate? Further if the answer to the above question is only the banking/Finance/insurance, then is the management also looking at mapping asset sub classes which it should offer to the target segment ( Retirement planning).
Hi Rajesh,
Thanks for the feedback.
XYZ can enter into a variety of sectors based on the current distribution of preferences among *broad* asset classes in the target population. There are some asset classes such as Gold, where XYZ can't do much but the info would be good to know that such % is gunning for Gold.
XYZ can enter the property sector not to compete with DLF or anything but perhaps in housing finance, if housing is seen as sufficiently attractive for a good % of the target population.
Point being that a lot of asset sub-class level detail would probably not be required for this study. XYZ will need to commission follow-up detailed studies if this one throws up interesting opportunities.
Warning: Longish post - but worthwhile read, IMHO.
Hi All,
There've been some queries that might be more widely shared by the class at large. Let me use this post to try to answer some.
Vikram writes:
Sir,My response:
We were discussing the project and have a query:
We want the consumer's requirement at retirement.
But we would also need to present him with the investment he needs to make corresponding to a return.
Will we get information about premiums we can use for this purpose ?
Hi Vikram,
Kindly look at the comments section of this thread
http://marketing-yogi.blogspot.com/2010/10/phase-i-submissions.html
Maybe some idea of what I have in mind might emerge. In any case, you don’t have to buy into my ideas on this one and can choose to go your own route.
I shall address your post separately and then postthe same on the blog.
Going forward, I'd rather questions of a general nature be taken to the blog (a response is guaranteed within a few hours at most!) where the entire class can access, share and benefit from the same.
In general, I've avoided giving much 'guidance' on the project for fear of biasing your natural thought processes and responses to the scope document. So let me stress, my ideas aren't necessarily the only 'right' approach to this problem. If your approach differs, fine.
People, please lookup the comments sections to different posts for relevant Q&A. Pls use the comments section for follow-up Q&A. Let me first quote what I wrote in the comments section of 1 post below:
Let me add to this now. Think from the Client's perspective also. The client is interested in a known-unknown - the broad range of asset classes currently (and prospectively) finding favor with the target population, and the current state of awareness and attitude among the target population towards these alternatives.
Respondents may be unable or unwilling to answer certain questions.
In such cases, we tone down the info content of the questions. For example, instead of asking folks to reveal their household incomes, we ask merely that they pick this off a list of ranges.
Similarly, IMO, since we are interested in the distribution of preferences across asset classes, a rough, relative measure of such allocation might suffice. E.g., "Out of 100% that I put into longterm savings and investments, I'd ideally go with 15% in equities, 15% in a private provident fund, 40% EMI on house loans, and the rest in bank FDs".
IMHO, respondents will hopefully give a true picture, on average, of such allocations. Down the line somewhere, one might casually inquire into current income p.m. range and % of that put into longterm savings.
Again, this is entirely my static view and you needn't follow it 100%. Based on your evolving plan, go with what you think is best for phase I.
Some folks may not even have thought about retirements and longterm savings. What is the % of such people in the target population? What markers - demographic or psychographic - may distinguish them from other folks?
Then again, some other folks may have thought deeply about this issue and planned their retirements in some detail. What is the % of such folks, their motivations, background, Identification markers, and importantly, their current portfolio like?
Some people would be in-between the two extremes above. How amenable are they to the message that retirement planning is important? How best to reach them via media exposure?
In general, how do people view investments such as property, gold, stocks, etc? Long term savings-vehicles? Inflation hedges? Retirement income sources? I don't know what else you may come up with. These things I am thinking up as I type.
Even more generally, what factors affect the awareness, attitudes, ideas and actions of people vis-a-vis the issue at hand? Is it their age, their stage in the life-cycle, their family size - # children, their employment type, peers and social circle, the influence their parents' choices in this regard had on them, opportunities at hand, etc etc.
If you want to use standard Marketing frameworks you have learnt elsewhere here, be my guest. For instance, I can think of the AIDA framework in Kotler that could be gainfully applied. AIDA stands for 'Awareness-Intention-Decision-Action' and measures where potential consumers are on this continuum vis-a-vis a particular product.Erratum: 'Decision' was previously incorrectly IDed as 'desire'
All these things may or maynot figure in your design of the questionnaire. All of them probably cannot be accommodated anyway in the spacetime constraints we are operating under. This then, is the challenge for you, the call you have to make.
The 15+ person-hours (at least, I expect) your group spends on the questionnaire design should be great leaning in the process of satisficing competing demands and requirements, optimizing available resources within set constraints and all in the framework of a complex issue at hand. This is the real learning the course hopes to impart - the bookish gyan is always there anyway.
Chalo. The post is long enough already. I should stop. Pls feel free to use the comments section for follow-up Q&A, feedback, critique and discussion
Sudhir
P.S. Also meant to mention some best practices from last year that came into handy. Some groups modularized the questionnaire and had subgroups of 2-3 work on separate modules. Apart from the efficiency of parallel-processing that this mechanism gave them, an added advantage was that one subgroup could pre-test the others' modules. Am not sure if this will work for your particular group or not, but something to consider.
Also, let me once again, strongly urge groups to setup and use free Wikis to coordinate activities across time and distance. Pls see the relevant blogpost below for the link to Wiki creation.
Saturday, October 23, 2010
New stuff up on Blackboard
Hi all,
Have putup 2 new items on blackboard - (1) Rcode and height-weight data for those among you who may want, for whatever reason, to replicate the simulation exercise we saw in class in lecture 4.
And (2) a sample homework solution that a student sent in and graciously agreed to share with the rest of the class. Now I think except for one question (the price/oz one) the rest of her work seems fine, but I may be wrong. Kindly go through, compare with your own answers and let me know if all is well. BTW, 'top-selling' referred to revenue based, unless otherwise specified.
Am going through this midterm preparation process. Its a thankless job but somebody has to do it. And invariably, I can't get rid of the nagging feeling that the exam is too easy. In any case, typically about 50% of the marks are the peaceful and straightforward variety. Another quarter moderately tough and the last quarter is tricky. Typically.
Sudhir
Have putup 2 new items on blackboard - (1) Rcode and height-weight data for those among you who may want, for whatever reason, to replicate the simulation exercise we saw in class in lecture 4.
And (2) a sample homework solution that a student sent in and graciously agreed to share with the rest of the class. Now I think except for one question (the price/oz one) the rest of her work seems fine, but I may be wrong. Kindly go through, compare with your own answers and let me know if all is well. BTW, 'top-selling' referred to revenue based, unless otherwise specified.
Am going through this midterm preparation process. Its a thankless job but somebody has to do it. And invariably, I can't get rid of the nagging feeling that the exam is too easy. In any case, typically about 50% of the marks are the peaceful and straightforward variety. Another quarter moderately tough and the last quarter is tricky. Typically.
Sudhir
Friday, October 22, 2010
Phase I submissions
Hi all,
The phase I deadline is midnight (or 11.59 pm) of wednesday 27-oct. It was originally intended to be 25-oct but group formation delays have pushed it slightly further up.
There are 2 deliverables in phase I. A working link of your websurvey must be entered into this google form:
link
Immediately after, send an email to your respective AA with a word doc soft copy of your questionnaire that should also have the URL on it somewhere.
Any doubts, queries, clarifications etc, pls use the comments section of this thread or email me.
Sudhir
The phase I deadline is midnight (or 11.59 pm) of wednesday 27-oct. It was originally intended to be 25-oct but group formation delays have pushed it slightly further up.
There are 2 deliverables in phase I. A working link of your websurvey must be entered into this google form:
link
Immediately after, send an email to your respective AA with a word doc soft copy of your questionnaire that should also have the URL on it somewhere.
Any doubts, queries, clarifications etc, pls use the comments section of this thread or email me.
Sudhir
Weekend Mailbag
Update: More Mailbag:
V writes:
Sudhir
Hi all,
Shall keep updating this post with more mail queries that are sharable with the class at large. So kindly check for updates every once in a while.
'S' writes:
V writes:
Could you please post sample questions for both parts of the exams?My response:
One more que: the MCQs on the slides are any representative (of content and level of difficulty) of the questions you might ask in the exam?
Hi V,
Am yet to make the main paper itself. The sample paper will likely have to wait. I hope to release a few sample questions at least 24-36 hrs prior to the exam.
The MCQs will be a mixed bag - some easy, some moderately easy and a select few, tricky. The class slides are more to stimulate discussion, critical thinking etc than to 'settle' an issue this way or that. Hence, no, the MCQs in the exam will leave a lot less room for interpretation and controversy, I hope.
Sudhir
Hi all,
Shall keep updating this post with more mail queries that are sharable with the class at large. So kindly check for updates every once in a while.
'S' writes:
Dear Professor,My response:
Would it be possible to please upload the answers to beer dataset practice questions. It will really help us in verifying if we have got the answers right.
Hi S,Sudhir
I haven't made any solutions to the beer dataset myself. However if you wish to share your solutions with the class (anonymously, if so preferred), I'll be most happy to do so on the blog so that other people can compare answers and confirm/disconfirm particular solutions.
Thursday, October 21, 2010
Exam announcement - the Coop case
Update:
Folks, Piyush has raised an important question regarding phase I submissions in the comments section to which I have replied there. Kindly carry that conversation forward in the comments section in case you have any further queries or comments to add. I shall later make a separate blogpost to provide more detailed info on this matter.
Hi all,
ASA confirmed today that you will all be permitted to take your course-packs into the exam hall for the entire duration of the exam. And with good reason. The chicken coop case in the course pack *might* see some action in the exam. Again, 'might' is the operative word here, no guarantees something from there will land up. But if I were you, I'd not take chances with skipping a detailed read-through of the case. Just saying, only.
Do let me know if you're planning to use Wikis to coordinate group project activity.
And pls ensure JMP is up and loaded and ready before mid-term.
Sudhir.
Folks, Piyush has raised an important question regarding phase I submissions in the comments section to which I have replied there. Kindly carry that conversation forward in the comments section in case you have any further queries or comments to add. I shall later make a separate blogpost to provide more detailed info on this matter.
Hi all,
ASA confirmed today that you will all be permitted to take your course-packs into the exam hall for the entire duration of the exam. And with good reason. The chicken coop case in the course pack *might* see some action in the exam. Again, 'might' is the operative word here, no guarantees something from there will land up. But if I were you, I'd not take chances with skipping a detailed read-through of the case. Just saying, only.
Do let me know if you're planning to use Wikis to coordinate group project activity.
And pls ensure JMP is up and loaded and ready before mid-term.
Sudhir.
Final group formation
Pls find the google spreadsheet for final group formation.
MKTR final group list
The allotment is *final*. We have tried to maintain expressed preferences as far as possible.
MKTR final group list
The allotment is *final*. We have tried to maintain expressed preferences as far as possible.
Ice-cream survey dataset & exercises
Hi all,
The AAs will soon upload a zipped file containing 4 files - (i) soft copy of the questionnaire (ii) practice questions word doc (iii) results summary sheet (iv) results raw data sheet.
Kindly ensure you can solve all the exercise questions. Feel free to email me with questions if any, or meet the AAs with the same.
Even though you already had the hard copy distributed in class, the soft copy of the questionnaire has been put there for a reason. I'd strongly urge each group to use it to practice translating a questionnaire into websurvey. Embed the SKIP logic into the websurvey while you are at it, on esurveyspro.com free account. Do it concurrently with the design and development of the Phase I project questionnaire. Use this weekend to its fullest.
Sudhir
The AAs will soon upload a zipped file containing 4 files - (i) soft copy of the questionnaire (ii) practice questions word doc (iii) results summary sheet (iv) results raw data sheet.
Kindly ensure you can solve all the exercise questions. Feel free to email me with questions if any, or meet the AAs with the same.
Even though you already had the hard copy distributed in class, the soft copy of the questionnaire has been put there for a reason. I'd strongly urge each group to use it to practice translating a questionnaire into websurvey. Embed the SKIP logic into the websurvey while you are at it, on esurveyspro.com free account. Do it concurrently with the design and development of the Phase I project questionnaire. Use this weekend to its fullest.
Sudhir
Quiz 1 for sections C & D
Folks,
We had a 10 minute quiz based on the beer dataset for sections C and D today.
There were complaints that some haven't any prior familiarity with Pivots etc and hence are at a natural disadvantage in such a quiz. I broadly agree with the caveat that I'd opened the dataset to the class and the associated homework quite some time ago, put up links and online tutorials to lookup functions and pivot tables here on the blog days ago, and had repeatedly hinted the same might make its way into a quiz or exam. I would expect folks to work these issues out at that time itself.
However, after a conversation with the AA and in the interest of fairness, we have decided to
(i) run a special second quiz on the same topic for people who score between 0-2 out of the 8 marks assigned for the first 2 questions.
This second special quiz will be tougher than the preceding one, mind you.
(ii) have a tutorial based on demand.
Hope that clarifies and helps. Going fwd however, let me clarify that anything homework related should be do-able by you. If you can't do it, ask your group mates or the AA for help.
I shall soon putup raw data for the Ice cream dataset and release a list of practice/homework/ exercise questions. Again, I cannot guarantee that the concepts asked won't landup in a quiz or exam somewhere. For your midterm anyway, you should ensure that all the datasets released are loaded in your laptop.
We had a 10 minute quiz based on the beer dataset for sections C and D today.
There were complaints that some haven't any prior familiarity with Pivots etc and hence are at a natural disadvantage in such a quiz. I broadly agree with the caveat that I'd opened the dataset to the class and the associated homework quite some time ago, put up links and online tutorials to lookup functions and pivot tables here on the blog days ago, and had repeatedly hinted the same might make its way into a quiz or exam. I would expect folks to work these issues out at that time itself.
However, after a conversation with the AA and in the interest of fairness, we have decided to
(i) run a special second quiz on the same topic for people who score between 0-2 out of the 8 marks assigned for the first 2 questions.
This second special quiz will be tougher than the preceding one, mind you.
(ii) have a tutorial based on demand.
Hope that clarifies and helps. Going fwd however, let me clarify that anything homework related should be do-able by you. If you can't do it, ask your group mates or the AA for help.
I shall soon putup raw data for the Ice cream dataset and release a list of practice/homework/ exercise questions. Again, I cannot guarantee that the concepts asked won't landup in a quiz or exam somewhere. For your midterm anyway, you should ensure that all the datasets released are loaded in your laptop.
Subscribe to:
Posts (Atom)