Social Data: Making Marketing Research Cheaper, Faster, and Stronger

Posted by: in Social Media Insights & Trends on November 1, 2013

There’s a chapter in the 1983 book Ogilvy On Advertising that I’ve always found amusing as an agency researcher. In this chapter, called “Jobs in Advertising”, Ogilvy goes through all of standard ad agency roles (i.e., planning, account, creative, media, etc.) and he systematically praises the virtues of each and the value that they bring to an agency and to clients. When it comes to agency researchers, however, he launches into a multi-page list of grievances about how research takes too long and how researchers favor precision and unnecessary levels of accuracy over quick, directionally correct, and actionable results. Thirty years later the grievances are, sadly, as relevant as ever.

If you’ve ever worked for or with an agency, you might already know a dirty secret about the research they carry out in advance of pitches or campaigns. The secret is that traditional focus group, survey, and panel-based research takes so long to carry out, synthesize and make sense of, these are often done in parallel with planning. If the research ends up supporting campaign ideas, the campaign is presented as if research drove it all along. If the research doesn’t support the campaign, it’s simply ignored.

This is all changing today. As the publicly available, historical treasure trove of conversational and behavioral online data grows, we’re relying less and less on slower, more expensive forms of marketing research. It’s taking my colleagues at WCG and me matter of days to come to derive the same types of insights that used to take weeks or months. And when we look beyond typical social media data use-cases, which are typically focused on social media marketing and community management, we’re identifying consumer segments, audience motivations/attitudes, and insights that inform a wide-range of product development, branding, creative, and other marketing activities that occur both on- and off-line.

Social Data: It’s Representational Now

The idea that content from social media networks might be a good replacement for marketing research isn’t a new one, but the general consensus among the research community has always been that it isn’t a very good one. A couple of years ago when Joan Lewis, the head of Consumer & Market Knowledge at P&G, announced to the Advertising Research Foundations’ ReThink 2011 conference that they’d be shifting some budget from survey to social media listening, the overwhelming response from the audience (and online) was: “you can’t do that because social media isn’t representative.” To paraphrase the non-believers, the folks who use social media are too young and too likely to only share very select parts of their day-to-day lives (i.e., really great experiences and product complaints). You can’t really understand the real marketplace by assessing the few anomalous people who are into the social media fad.

The “not representational” argument worked a few years ago when social networks were dominated by the young and social media-curious. But those arguments fall flat today. There is ample evidence that you can find just about any group sharing information about just about any topic online in 2013, including these statistics:

  1. 52% of Americans are actively using social networks (eMarketer, 2013)
  2. Adoption has increased across all major demographic groups; even 43% of Americans over 65 who are online, use social networking sites today (Pew Research Center, 2013)
  3. 24% percent of people who use social media, globally, report that they share “everything” or “most things” online (KPCB, 2013)
  4. The distinction between public and private is blurring for the next generation of consumers, meaning they actually will share anything.  53% of teens posted their email address online in 2012 and 20% shared their cell phone number (up from 29% and 2%, respectively, in 2006; Pew Research Center, 2012).
  5. Social media isn’t just a place to crack jokes about the Kardashians. Even when conversational analytics fail to deliver meaningful insights for a product or brand, people’s behavioral data is often very rich. For example, American’s on Facebook like an average of 70 pages (Socialbakers, 2013).

The Social Data Advantage

When you consider that social data is often representative, behavioral (not just conversational) in nature, often candid and can be sampled at the point of experience via mobile devices instead of retrospectively through questionnaires, there’s good reason to believe that it can serve the same purpose as traditional market research. When you add in the benefits of using social data, including the relatively low data collection costs, the ability to analyze the data quickly/automatically, and real-time data flow, I expect that it will disrupt the way market research is done in the near future.

Over the past 5 years, I’ve often thought that the results of social media studies worked best in conjunction with other forms of market research. If two or three separate methods of research produced the same results, I figured that the findings were meaningful and worth acting on. As I accumulate cases where survey research and social data research produce the same types of results, I’m less convinced that it’s always worth the time and budget to use surveys and panels and simply augment those with social data. At least once a month, I’m hearing clients say, “what you just found in one week usually to take us 3 months to discover.” There’s no defensible reason to spend $500k on focus groups and surveys that identify the same groups of consumers you could have identified at a fraction of the cost using social media affinity data.

What this means is that companies that harness social data to answer big questions, not just those related to how you should grow your Twitter or Facebook follower base, will have a huge advantage over those that don’t. For one, they’ll be saving hundreds of thousands, maybe millions, of dollars each year in research budgets. More importantly, they’ll be much faster at responding to changes in the marketplace since they will have insight into changing trends months, or even years, ahead of the competition.

The abundance of relevant social media data makes for exciting times in the marketing research world. We’re now at that place where data and insights aren’t just a convenient agency add-on when they happen to fit with an a priori strategic vision. Research can move at a fast pace that puts insights ahead of big decisions in every aspect of our work.

By: Seth Duncan

Find me on: Twitter
Pre-Commerce Check out W2O Group President Bob Pearson's new book, Pre-Commerce, in which he shares ideas for leaders to engage directly with customers to shape their brand and marketplace success. Now available for order on Amazon.com! http://amzn.to/bAmvFN. Join the conversation #precommerce.

5 Responses

Stay in touch with the conversation, subscribe to the RSS feed for comments on this post.

  1. Anonymous said

    Harnessing social media for market research seems obvious today, but there’s another important advantage – it’s dynamic. Traditional market research is a (detailed) snapshot in time and may very well be obsolete in a few weeks. Modern agencies and companies with internal marketing need to keep their eyes on the market 24/7/365.

  2. I find myself torn by this post. As an advocate for brands
    using social media in all areas of business, from customer service to
    competitive intelligence and product development, I agree that valuable information can be obtained using social data that sometimes might not be obtainable via more traditional research methods. However, as a market researcher with over a decade of experience in survey research, I think that it is still too early to eliminate the more traditional data collection methods altogether.
    Again, I am not arguing with your main point that social data is valuable, sometimes more valuable than data collected via traditional research methods. What I disagree with is your statement that it is representative. Just the fact that only 52% of Americans are actively using social networks shows that there is going to be bias in the data. Is it good enough given the cost savings? Maybe some of the time. But, if the decision being made can cost a company a lot of money, I think traditional research methods still need to be included.

  3. I’m getting your book!

    We (Mattr) obviously feel the same way. Interestingly, I think social data analysis, especially the qualitative like we do, may be more representational than surveys or focus groups. Here’s why:

    1) totally passive; no biases which inescapably accompany social or online surveys and in-person modes.

    2) business is converted and transacted via social media. There’s no call to action in a survey or a focus group – certainly not one with the scale of social media. This means that you can sell directly to the very large group of segments who you’re analyzing.

    3) no self-idealization in social. When you participate in a survey or focus group, it’s easy to project yourself as someone you’re not. Solid social media analytics can see past the context of what you write to your true personality. There’s excellent research on this topic.

  4. I think I’m with you. There are certain research needs, at different times of a product’s life, that will require traditional MR. Kevin Systrom said it best last week: “products in their early stages need hunch-driven design; it’s only after they’ve matured and reached scale that data-driven design is necessary to optimize it.”

  5. A point I left out of my earlier comment – definitely!

Some HTML is OK

(required)

(required, but never shared)

or, reply to this post via trackback.