Howard Moskowitz – I’d Like to Teach the World to Think

[Reposted from the Next Gen Market Research Blog]

Q&A with a Continuous Innovator and True NGMR 1%er Howard Moskowitz

 Last week I talked about ‘Outlaw Market Researchers’, the 1% who can’t help but innovate, BUT do so in a smart way because they understand proper research methodology inside out.

Today I’m happy to have a true NGMR 1%er with us, Howard Moskowitz Ph.D. I’m sure you’re already familiar with Howard and his many accomplishments in our field.

I met Howard for breakfast recently to discuss conjoint, text analytics and NGMR, and our discussion spilled over into the following post about his thoughts on our industry and more recent innovation and effort into taking Conjoint/Mind Genomics mainstream.

 

THCA: Howard, you’ve had quite the career in marketing research with some very notable innovations.

Some of these are methodological, like the use of response-surface methods and reverse engineering for product design, the use of conjoint methods for package design, and of course horizontal segmentation that Malcolm Gladwell refers to as a ‘game changer’ in the food industry, giving rise to dozens of sku’s and the paradox of choice. Also, you’re responsible for a number of major brands, such as Prego, Tropicana Grovestand Orange Juice, Diet Dr. Pepper Vanilla, and others. This is a lot of good work in the 42 years you ‘admit’ to being in the field. Thank you for this interview, and also for the fact that, as you told me, you’re theme is now ‘ I’d like to teach the world to think.’

Beyond, smarts, and excellent education and hard work, can you tell us a bit about why you think you’ve been so successful?

 

HM: Thank you Tom. Why successful? H’mm. Well, one thing is that I focused on what I believed in. That sounds so idealistic, but it’s not. I got into the field in 1974 through my work with Pepsi. I didn’t come into the field to make money. I came in because my late father of blessed memory, Moses Moskowitz, asked me ‘Howard..do you just want to be a scientist for the rest of your life?’ That made me realize I should apply my work, my science. Not make money, but apply the gift of science and knowledge that G-d gave me. I know that you were looking for a savvy answer. But it was simply ‘apply the science you learned to business problems.’ Oh money is good. But it’s not really the driver. To validate onself, that’s the real reason.

For your readers: I’m writing my memoirs to record the incidents of my life, incidents of education, family, and work. The memoirs are a work in progress. They are not PR about me, nor my own ‘humble boasts.’

https://www.dropbox.com/sh/6llnitg4ce2kmlj/AAC-n68FecoJlfXA70N10GO2a?dl=0

 

THCA: You’ve been in this industry for some time. Can you give us some thoughts on how you think the industry have changed for the better?

 

HM: First, by way of disclosure … I’ve been a vendor, supplier, whatever you call it for 42 years, since November 1, 1975. I’ve only worked on the supplier side. So what I tell you, good or bad, is from the viewpoint of a supplier.

The industry has gotten better because more people recognize the value of knowing what the consumer wants. Years ago, we often saw the marketing staff override the research results, saying that the marketers ‘just knew’ what to do, what the consumer wanted. Today researchers have won the respect of marketers.

The industry has also gotten better because the tools it uses are in the hands of many people. There is more thinking going on, more recognition of the customer journey, more tools available, and more flexibility to try new things.

 

THCA: How has it changed for the worse?

 

HM: There are a lot of things that have not been so good. I’ll list what I think to be wrong

I’m not going to be nice and diplomatic. I’m 73 now now, s, I have a low Net Present Value. That allows me to tell the truth, without worrying about the repercussions.

1. Narrowed Scope and Overly Specialized: Researchers do research, rather than get involved in the holistic issue of ‘why.’ We used to call this a ‘seat at the table.’ Today the technical sophistication is so great that marketers are loathe to involve the researcher in the deeper business problem. The researcher now positions himself/herself as a technical, data specialist. This isn’t just the case with consumer research. It happens in science as well. The scope of one’s ideas and vision has narrowed considerably over the years as the array of methods and topics has increased.

 

1. The researchers are often very young and inexperienced. They don’t have the perspective that they need. However, it’s not their fault. The competition is fierce. The researchers need to survive, and to survive they must spend more time than appropriate selling. They don’t have time to think. Furthermore, these young researchers are looking for an edge to let them get an appointment with the client. It’s natural to gravitate towards the glitzy and the new, because that gives them something to talk about.

 

1. Procurement, approved vendors, and ‘strategic partnerships, have destroyed thinking. The idea is to get a locked-in contract, a preferred status, a repeating project. The focus is getting in, not getting it done right. Getting on the client list, not solving the client problem, is the real achievement. At first one might think one’s getting a ‘good deal,’ locking in the supplier. One ends up with the A team selling, the B team presenting, the C team doing the research, and the top management having lunch to smooth over problems of under-delivery. That’s not to say that standardizing isn’t a good thing .. but in the words of my late professor from Harvard, SS Stevens, ‘standardization is the last refuge of the unimaginative.’

 

THCA: Other than Conjoint, which areas of research interest you most and least?

 

HM: When I started in 1975 I was only interested in products. Over the years, I became interested in whatever the client was offering as a possible research project. I can’t really say that within the research community I was particularly passionate about anything, such as audience measurement. It was the focus on surviving, on making a living, on getting that NEXT project, which was most interesting. Sorry for the blunt answer, the type of answer one might get in a psychotherapy session, when asked the question by one’s therapist.

In more recent years I have had the time to reflect on what REALLY interested me. I think that I was most interested in discovering the way people think, the specific ‘hot buttons’ of taste, of visual communication (packages, messaging), and of advertising communication (concepts, advertising itself).

But then the real question is  WHY the lack of a profound interest in something until the later years, when I was seasoned, tired, and realistic. I attribute that to the need to survive. The ‘something’ was survival, and validation of oneself.

I know I haven’t answer the question in the way that you meant me to answer. And you know that I have written many, many books, articles, chapters, and conference proceedings on different topics. None of them can I say interested me much beyond the moment. Each was fascinating in the extreme at the moment of doing it, or I could never have done it. But each was momentary, as if it were the only thing that minute, a shooting star for an instant, lighting up my sky, but disappearing entirely afterwards.

 

THCA: So let’s talk a bit about Conjoint. I know you’re thinking about Conjoint in some unique ways from our recent breakfast together when you were interested in increasing the scope tremendously, and possibly even merging of text analytics with conjoint.

For those less familiar with what you’ve been working on regarding Conjoint and BigMind, could you give us the quick elevator pitch?

 

HM: Please allow me first to tell my story about conjoint, and then answer your question. As you will see from my story, there is a profound, karmic aspect to it all. I’m a believer in karma and the need to answer one’s call

Everyone has a story. Here is mine, and I’ sticking to it!!! I was a first-year graduate at Harvard’s Dept. of Psychology. It was an early December day. I was sitting in the graduate student’s lounge on the 6th floor of Wm James Hall, the lounge adjoining the library, and the unofficial second home of the first year graduate students. Larry (Lawrence) Erlbaum walked in, then a detail man for Academic Press, now the retired Mr. Psychology who started Lawrence Erlbaum Press, and published many psychology books. We talked, and he gave me volume 1, number 1 of the new journal, Journal of Mathematical Psychology. I was interested in Math Psychology. The first paper was Conjoint Measurement: A new form of fundamental measurement, by R. Duncan Luce and John Tukey. I read it, and was hooked. Sadly, however, I would not be allowed to use the approach for my thesis, which was all well and good anyway because in 1965 conjoint analysis was unwieldly. It became better, more tractable, when business researchers Paul Green and Jerry Wind of Wharton used it for applications.

 And now here I come along, with a lifetime of conjoint, or at least since 1980, when I began using it in the commercial world. To me conjoint was the mixing/matching of ideas according to experimental design, presenting the combinations (vignettes, concepts, whatever you call them), and deconstructing the reaction to the contribution of the components.

Most important for me … make the stimuli interesting (phrases that a human being can understand, not just sterile words), combine them, use regression analysis to see what works, do this at the individual level, and then see whether there are clusters of people with similar patterns of how they react to these meaningful phrases.

I guess I differ from most of the other practitioners of conjoint, who focus in on optimizing the bundle of attributes of a product, using whatever modeling tools are thought to be most powerful. I have always been more simple. First, I used simple ordinary least-squares, and dummy variable modeling, which means that the independent variables are 0/1, absent/present, and the statistics show how many points are contributed by each element. And second, I focus on messaging, on what do you say, with the variables I use being messages, or pieces of a picture.

 

THCA: So first let me say I’m very impressed with your scale of vision here, and I’m certain you are right that this could be hugely useful in solving a lot of problems. That said I have a few questions and concerns about Conjoint that maybe you can help me and the readers understand better.

It seems that there is a small but very passionate group of researchers that have, pardon the analogy, are really drunk on the conjoint cool aid. I’m talking for instance about the AMA ART Forum guys etc. I assume you’re familiar with this dare I say ‘geeky’ but extremely passionate bunch who have created an entire event about this seemingly complex technique. Yet outside of that relatively small group, rather few researchers seem to know much about it or feel very comfortable discussing or implementing it. Why do you think this is?

 

HM: Please excuse me for what I am going to say. I don’t think that they are in love with the problem they are solving. They seem to be more in love with their creation of a method to solve the problem. Sort of Pygmalion the sculptor falling in love with his sculpted statue of Galatea, or for others not so steeped in Pygmalion, perhaps My Fair Lady.

Why do I say these infuriating, incendiary words? Simply because there is no corpus of literature using conjoint measurement to solve a series of related problems, except for perhaps my new 18 volume series, the New Novum Organum, using Mind Genomics, my version, to erect an empirical science of experience. Here is the link:  https://www.dropbox.com/sh/v7sobapniifvbep/AAB-Tn6z6uLQDZQSN0UDFriya?dl=0

Most of what we read in the conjoint literature, and hear at the meeting, deals with the method, some new aspect, with the substantive problem being simply an afterthought. Well, when the problem is an afterthought, a substrate into which one implants this specific, new, pay-attention-to-me conjoint method, the outcome is determined. Conjoint, and it methodological niceties, become the focus, not the problem it is going to solve. And thus, the ART Forum has evolved to what it is, a showcase of method, rather than a more substantive-topic of real-world issues, applications, knowledge, and solutions.

One should also realize that the research community, unlike the marketing community, prides itself on two things; personal ‘brains’ (the Hero’s Journey in the words of Joseph Campbell) and the contribution to the company. The abstruseness of many variations of conjoint are as much to proclaim the researcher as they are to solve the problem. This is probably going to arouse the ire of many professionals in the field, but the reality is that I personally, at the age of 73, with a 48-year professional history, with many studies using conjoint, can barely understand the methods, nor understand what I am being delivered. And that failure to understand has grown over the years, not diminished. That’s my story .. and I’m sticking with it!!

 

THCA: My first serious exposure to training on Conjoint techniques was when taking Burke Institutes 3-day seminar on it. I’ve got to say I think they have some of the best training courses on almost any subject in traditional marketing research. Anyway, based on that training, and subsequent experience, it seemed to me that a major limitation to conjoint is that there are only so many ways it can be tweaked to be made more efficient (Sawtooth and others have been working on this for some time/Adaptive Choice Based etc.), but the technique is by its nature extremely taxing on the respondents with multiple levels that need to be rated even when using adaptive techniques.

This while Pew Research institute is arguing that survey response rates continue to drop and soon there literally won’t be anyone dumb enough to want to take a survey. Would love your thoughts on this seeming dilemma?

 

HM: Oy. That’s my first reaction. And in that, I’ll tell again a story. I was a young scientist, 24, just starting my work at Natick Laboratories. It was March 1 or so, 1969, almost a half century ago. My respondents were soldiers. I was doing ‘taste mixtures’ research .. what is the taste of a mixture of say sugar and salt. We had 48 samples to work with. Each panelist (soldier) had to do the 48 samples. They were terribly fatigued after the 10th sample. Couldn’t taste. What was I going to do? I asked the Laboratories to give me $10 to fund a 4 hour session, i.e., $10/person for four hours, in 1969 dollars. Guess what …. No fatigue, happy non-military residents, chirping away. I wrote about the ‘magic’ of paying respondents. And in my writings I asked ‘what other business wants to get their “raw material” free? The “raw material,” of course, being the respondents.  I presented this at various companies. One company in particular, in CT, had a research director, JB, who jumped up and said ‘ Howard, by paying respondents … YOU are destroying the research industry.” Well, here we are. No pay, no respondent. Homo Economicus wins, not Homo Emotionalis. Your opinion may count, but I can’t get it from you with only sweet talk. No how, no way, no longer.

I created conjoint methods which allowed people to do subsets of elements. I wouldn’t create an individual model, but rather pooled the data. Later, I created IdeaMap, which in its early versions could handle hundreds of elements, estimating for each person the coefficient or utility of untested elements. Later, I migrated to the web, and to individual-level models, forcing me down to a maximum of 36 elements. And in my student-version (below), I reduced the 36 elements to 16 elements in 24 vignettes.

 

THCA: I know you’re currently working with new Israeli developers in getting your software to market. In the past I haven’t had the best luck with offshoring, even simpler tasks than development. That said I realize Israel is leading in this area compared to many other countries. Could you tell us a little bit about this experience so far, and any tips you can pass on?

 

HM: Thank you for that lead-in. My current IdeaMap.Net software is about 20 years old. I keep getting hints to update it, but I’m not going to, at least not now. The software works, has been battle-tested, has produced big successes and has allowed me to write 20 books and untold articles on various topics, using Conjoint Measurement and Mind Genomics, as well as helping a lot of clients and value-add resellers.

But I am working in Israel for a different reason. As I keep stressing in this interview, I’m no longer young. I’m no longer interested in WiiFM, what’s in it for me! I want to give back. G-d gave me a good technology, energy, enthusiasm, and resilience. I often wondered why. I found that ‘answer’ when my slightly younger sister (5 years younger) and my two sons, both of them, 29 and 34 years younger than me, respectively, told me: “ Maybe you should create a version for young people, whom you teach to think creatively, and enjoy experiential learning.”

And so I’m putting Mind Genomics into a free APP. Anyone can do a small-scale experiment, and get the results published on a website. If companies want to buy the data, the company pays $200 for the study. The ‘researcher,’ typically a young person gets about half of that!!

Imagine millions of young people learning to think, using a student-version of IdeaMap (our version of conjoint analysis), getting the results, having the data available free for download (total panel and 2 mind-set segments), and being able to use the data for projects, for interviews with prospective employers, as a way of having fun, funding clubs, or going to local businesses and acting as junior consultants. Or…have companies use young, inexperienced researchers, to ‘crowd source’ and test ideas.

The final vision of this effort, APP and database, we call BIG MIND. The BIG MIND database can be populated from around the world with small-scale, affordable ‘projects,’ small studies of maybe 50 people, and 16 elements. What happens when we have 10 million or 100 million people ‘learning how to think,’ showing their stuff intellectually, contributing to a world class database, a true experiment-based ‘Wiki of the Mind’

Here is our vision, and material, of a non-sales nature, to share with your readers

https://www.dropbox.com/sh/yuamdn1c3xnesya/AAArCtqrDsKXvhN9vdjtWOxua?dl=0

 

THCA: Stepping away from methodology and technology for a bit, on a more human/psychological level, can you tell us a bit about what you’ve learned about communicating your ideas to other researchers and perhaps more importantly clients? What is the key to getting a new methodology or product idea accepted?

 

HM:

 

1. Education: I tried to show how the approaches work. I failed.

 

1. Proof: I tried to make the clients some money, which I did. But, as the bible says, and those who know me hear me quote, almost ad nauseum, “There arose a Pharaoh who knew not Joseph.’ You’re only as good as your last job.

 

1. Sophistication: I tried to advance the field, and did so, but the field went back to simple

 

1. Candide: I left the field, became a writer, a scientist, and now with Big Mind, I’m helping the people. I am ‘cultivating my own garden.’

 

 

THCA: What tips would you give researchers who want to make a mark like you have already done?

 

HM: The real opportunity for tomorrow: Look to make knowledge a B2C effort , business-to-consumer.t. Today’s world is changing. The day of the expert who hordes knowledge is over. Share knowledge, following the notion of Big Mind. Use knowledge to help many people, not just the few. In some ways, the Big Mind and its sister effort, Vox Populi (for governance) are my personal B2 C efforts, where reach and touch the people of the world, not the researchers of the world. It’s new, it’s exciting, and after a lifetime to B2 B, business-to-business, it’s where I want to make my stand, where I want to rest my soul..

Many thanks to Howard for sharing with us on the NGMR blog today. I’d love to hear your thoughts or questions in comments section below.

 

@TomHCAnderson