By Tony Cosentino
After reading a recent article from Lenny, I was hit by the idea that perhaps market research doesn’t have a structural problem as much as it has a perceptual problem.
I’ll give you a couple of examples of what I mean. I went to OracleWorld recently and listened to Larry Ellison position the idea of engineered systems and cloud computing. Of course, this is what Sun Microsystems founder Scott McNealy (who also founded social polling company Wayin) said a long time ago. According to McNealy, the reason things didn’t work out for him and Sun was merely because he didn’t call it cloud. Ironically, Oracle acquired them and is now positioning themselves as a big-data and cloud company.
This is also exemplified by the current funding environment both in terms of the startups in Silicon Valley as well as publicly traded companies. Companies like MuSigma and Absolute Data have raised tens of millions of dollars. I would argue that these companies are essentially market research firms, but they happen to be reading from a McKinsey playbook rather than a data collection playbook. I was speaking with an executive the other day of a publicly traded company and asked him why his company went up almost 10% that day while the rest of the market was down. He laughed and said that an industry analyst had initiated coverage and called them a “big data” play.
I think it may be helpful to share a little of what I am seeing in IT right now, since I believe these trends are a primary driver of what is occurring in market research. The cloud is taking shape (no pun intended) and this is forcing companies to talk less about the size of their server, and talk more about the size of their information competitive advantage. The incredible innovation is forcing companies to take an outward-in approach by focusing on the business use case, the data they need to prove that case, and then working on the technological stack from there. Prior to this time, the CIO really called the shots and the technology stack was the critical piece. The shared CIO/CMO agenda is to tie together web analytics, traditional marketing analytics (including attitudinal), and enterprise performance management data. At the core, it’s about tying together behavioral data with attitudinal data and making sense of it. (This latter theme should sound very familiar to those in the market research industry.) The big difference on the business end of this equation is the proliferation of promotional channels as well as consumption channels. On the IT side, data models based on NoSQL as well as in-database approaches are taking the processing (read: analytic modeling) directly to the data and allowing us to get away from flat file driven models into true multi-dimensional space. This in turn allows us to look at how the real world works. The technology is changing the business so much that SAP is essentially blowing up their current approach and pursuing an innovation-based developer angle and funneling billions of dollars into creating a new eco-system around big data. This approach includes a half a billion in venture dollars as well as hundreds of millions of dollars in channel and customer investment. The point is that while all of this technological change is occurring, the biggest struggle is, and will continue to be the issue of the last mile; that is how the data will be disseminated and consumed within the organization and how it will support decision making.
In a way, this challenge is the same one that the market research industry has been facing for many years. If you complete a great piece of research, but it sits in a binder on the shelf and collects dust, it is essentially worthless. The same is now true in the technology industry and in business at large. You can invest millions and even billions of dollars to integrate your information and analyze it, but if nothing is done with the output, it is useless. To me research firms can help with this sort of last mile challenge.
The biggest part of today’s skills gap in the market is the basic modeling and analytics skill set. McKinsey wrote about this earlier and I followed this up in my own blog post, entitled, addressing the analytics skills gap. To me, this is where the market research industry really has the ability to shine. The market research industry is of course full of good thinkers that know data and often have very good modeling skill sets; we know the difference in data types, quality control processes and data cleansing, information transformation and integration; most of all, we know about experimental design, probability, and statistics. These are skills that are severely lacking in business and the demand for these skills will only go higher and higher. The difference now is that we’re doing these things not only on small data sets, but on massive data sets.
It’s hard to argue that the market research industry does not have structural issues as well. Certainly the age of designed data, as Census Director Grove calls it, is coming to an end and some of the large market research firms are sitting on a pile of underperforming assets. But I would argue for the firms that are not under the same constraints, the issue is more of a perceptual one. So I ask the question: is it simply that the market research industry suffers from bad branding? Branding is one of the areas where many market research firms have given much advice over the years. Unfortunately, it’s often the shoe-makers children that go barefoot.