Transformations in Next-Generation Research

One of the best parts of my job is that I get to spend significant time with Communispace clients. My goal is to meet with at least 50 clients in the course of a year, and when I see them, I often ask these questions: What have you learned since we last met? What is changing in your field? What did you used to think was true that no longer is?

The answers are always fascinating—and recently, we synthesized comments from 100 insights professionals about their “New Normal.” The result is a list of 8 New Rules for Listening to Customers. As follow-up, we’ve gotten reactions from another 200 people, and we are currently getting additional responses from attendees at the recent ARF Re:think Conference in NYC where I’ll be presenting these to a large audience for the first time.

#1: Manage Your Social Media Jitters. In these days of Web 2.0 and conversations, it’s a constant that some consumers will tell you that your data is just plain wrong. Sometimes they are right, as in the case of members complaining last year about changes in Facebook’s privacy policy (which Facebook subsequently changed back). However, sometimes their truth is not reality for others. Example: in last year’s Motrin Moms ad “disaster,” we found that the majority of online moms had never even heard about the problem, and most of them thought that the ad was just fine. We need to listen hard to consumers, for sure, but if your target consumers are not on Twitter, for instance, then you don’t necessarily need to change strategy because of an angry tweet or two. Market researchers told us that their internal customers were just plain jittery and that they needed to be ever more educated about social media in this new age of customer engagement.

#2: Game-Changing Insights Don’t Usually Come From Testing. Time and again, clients bemoan the enormous investment they make in research that proves what they already know. This doesn’t mean that testing is unnecessary, but it does mean that equal emphasis needs to be placed on more open-ended discovery—leading to unmet needs, competitive white spaces, and new texture that can transform organizations. The most recent example of this is the Leno/Conan debacle, in which the data seemed to work, but where NBC completely missed how a generation of late-night TV watchers like to see the news, laugh, and go to sleep—in that order.

#3: Go Beyond the Ad-Hoc-Ness of Research. The notion that every time clients have a question or project they need to recruit a new group of consumers is expensive, slow, and impractical. The result of ad-hoc projects is a constant focus on whom to ask and how to recruit them, with little room left to focus on finding insights. Although many thought that the true value of online research would be related to cost-savings, the bigger benefit is that the web enables you to have continuous conversations with customers. Thus, the focus group gets extended (in case consumers have other thoughts or change their minds), the survey can reveal the “what” and the “why,” people can change their answers, and you have a chance to understand more than a small slice of your customers’ lives. This is reinforced by our new research on post-recession consumers, conduced with our partners from Ogilvy, in which a series of ongoing conversations with consumers gave texture about their lives that we could never have gotten from one project. This new research is available here.

#4: Cutting Edge Technology + Poor Research Methodology = Lousy Research. Just because a technology is new doesn’t mean it will result in the kinds of breakthroughs clients are looking for. Clients told us they were becoming more skeptical about “the shiny new thing” in and of itself—especially when it didn’t help accomplish the goal. Bernie Malinoff of Element 54 has done fascinating research to support this issue, showing how new fangled Web 2.0 surveys are often more difficult and less valid. The days of saying, “Let me see your software demo” will be replaced by questions about the fundamentals: How do you get people to open up? How do you create trust online? How do you learn what people really think?

#5: Don’t Mistake Data for Insight. In 1995, a financial services client told me he estimated that he could fill 12 Manhattan skyscrapers with data that his organization didn’t use. That was before the Internet Age, and thus I’d imagine they could now fill several city blocks. It’s just not true that whoever has the most data wins. Executives are hungry for “something I don’t know,” “something that will change how I think about my business,” or the like. We need to synthesize and report our findings in a way that gets to the core issues. The 70-slide PowerPoint presentation just won’t get read because the author can’t net it out.

#6: Top Executives Would Rather Have Fast Than Perfect. When executives have a burning question, the response in 2010 can no longer be, “If you give me $35,000 and 6 weeks, I’ll come back with a binder full of statistically significant information that will knock your socks off.” Too little, too late. The response that works is, “I can’t give you a perfect answer, but how about if I give you an early read from 50 consumers tomorrow morning?” That kind of response is in synch with the urgent pace of business today—and it’s more likely to add value to the process.

#7: Don’t Underestimate the Power of n=1. This is my favorite client quote, told to us recently by our first client, Tom Brailsford of Hallmark. Tom has several degrees in statistics and math, and he knows that there is safety in numbers. Still Tom is struck by how the big breakthroughs often happen away from the spotlight. His quest these days is often for something different than a pie chart: the lone voice that just makes him say, “Hmmmm…” This is substantiated by scores of clients who have told us that one emotional verbatim can move mountains in the executive suite.

#8: The Future Engagement Will Trump Sample Size. In an age when everyone is a professional respondent, the new currency will be less related to how many people receive the survey or how many consumers are in the online community. Instead, the metrics will relate to engagement: How many people actually participate? How much time will they spend? Will they do more than fill out a quick poll? How honest will they be? How hard will they work to help us understand their lives, dreams, and frustrations? Will they be engaged even if we don’t pay them much? The answers to these are the holy grail of 21st century listening.

I’m interested in your reactions: what you like most, what you like least, and what you think is missing. Post them here, or email me at, and I will send everyone’s thoughts out in the next month.


9 thoughts on “Transformations in Next-Generation Research

  1. I’ve always been baffled as to why the kind of innovation in marketing research, as mentioned above, did not come from MR firms in the first place. The MR industry largely ignored anything to do with social media and online communities until late 2008 when the industry saw its growth hit a wall. No wonder, as Coke recently stated, 80% of MR budgets is spent looking backwards, a reflection of the legacy MR culture of selling ‘research by the pound’. By contrast, online community insight firms have been operating in the ‘here and now’ since their origins in the early 2000’s, using an ‘outside-in’ approach to applied marketing without undue emphasis on research ‘technique’.

  2. I like point #7, “Don’t underestimate power of n=1” but probably not in the same way that the client intended.

    In text analytics we can now conduct a census of what is being said in a discussion board etc. So we don’t have to rely on sampling. So while our sample in a given month might be n=5000 that 1 person who said something differently is no longer an “outlier” but may be saying something really important.

    We have to catch what n=1 says, they may be alerting us of a new opportunity or a serious problem which if not nipped in the bud can grow.

    So even in n=10,000 pay attention to each single person, n=1

  3. Ted has a great point. Isn’t it interesting, though, how the major players in any industry find it difficult to be the innovators? I think that it’s hard to let go of the business model that works for you, the core competencies you have built, and so on. Look at newspapers. You know what they should do, but how to get there?

  4. #3 I really like your point highlighting the problem that so much of the energy (and money) going into ad-hoc projects is spent on recruitment, to the detriment of the actual point of the research. The fact is that panel recruitment for “traditional” research is an incredibly difficult process, and it’s gotten steadily harder and harder over the last 20 years. Having said that, I do believe there will always be a real requirement for ad hoc surveys, e.g. for pre and post awareness testing, where you need entirely different samples. The good news for ad-hoc, is that I am sure that intelligent, innovative use of web capabilities will help to introduce time and cost savings to the process.

    Meanwhile, even for traditional panels (e.g. TV Ratings or consumer CPG panels) there are usually fixed time-limits on how long a panel member can remain on it. This is because of the belief that by virtue of being on the panel for a long period of time, panellists’ views and/or behaviour may change. Have you studied the implications of this for panel members of online communities? I can understand the value of an ongoing dialogue, but I feel that these panellists would be especially sophisticated, and perhaps even more open to changes in their views/behaviour than traditional panellists.

    #5 Don’t Mistake Data for Insight. Let’s beware the danger of experiencing a “new normal” of senior executives believing that, with all of these great new technological advances, it is now easy to produce Insight. Technology can certainly help to speed up research; to reduce costs; to significantly expand on the depth of information provided by respondents; and to use sophisticated modelling techniques to pinpoint fluctuations or new trends. However, to nit-pick Insight of mountains of Data, we still need lots of the same, old-fashioned resource: intelligent, motivated and really well-trained individuals, who can take time to mull over what are the really key points of relevance to their clients’ market/brand/situation. And ultimately, the best Insights occur when the research practitioner has an excellent, ongoing dialogue with the client, and fully understands their business priorities and issues – it’s still a two-way street.

    #6 Haven’t Top Executives ALWAYS Wanted Fast Over Perfect? It just hasn’t been possible until now. There have always been two opposing forces at work in research: the client/boss wanting the answers NOW, and the research practitioner wanting to use the best methodology to generate those answers. For the incredible new research tools such as online communities to be harnessed effectively, we must:
    · Continue to establish high quality research briefs from the client (perhaps one of the most overlooked aspects of market research)
    · And continue to press for the application of high quality research techniques. Yes, we have amazing new tools, but they still require skilled handling (let’s not throw the baby out with the bath water).

  5. Annabel, to answer your questions:

    Re the effects of keeping consumers for long periods of time, here’s one report on it from our research team:

    Re your comments on item #5, I completely agree! Despite all of our text analytics tools, there is nothing that results in insight better than people who have the ability to synthesize, analyze or find those “nuggets”. You just can’t (yet) automate that.

    Re your comments on item #6, I think executives now understand that fast is often “an early read”, or a way to check on a “hypothesis”, and we need to loosen up and be OK with that rather than assuming that executives will be irresponsible with the data.

    Thanks for the perspective!

Leave a Reply

Your email address will not be published. Required fields are marked *