(This is the text of a talk I gave at the ARF ReThink 2009 conference in March 2009. It was part of an “Inspiration” session run by Kantar, and was the second part of a part on respondent engagement – the first part focused on web surveys. I will add pretty pictures when I get a faster connection!)


To understand how social media changes participant engagement in market research you need to understand one crucial fact: that research data can now be a social object.

What’s a social object? The idea was developed a few years ago by Jyri Lindstrom, who founded jaiku.com. Lindstrom’s theory says in essence that relationships in social networks aren’t built around interpersonal connection so much as connections via social objects. In layman’s terms, a social object is something people talk about.

And the web turns out to be brilliant at creating these. What’s more, it creates layers of social objects, which in turn allows different layers of engagement with a site. Take YouTube, for instance. You can use YouTube as a passive consumer, simply watching the videos. But you can also rate the videos, or comment on them. You can upload your own. You can look at channels, or befriend other uploaders and subscribe to their videos. You can organise videos into playlists, and then rate and discuss those.

Layer upon layer of social objects, layer upon layer of potential engagement.

Let’s apply that thinking to research. The individual’s response is an obvious social object – already we’re seeing techniques like crowdsourcing become popular, where users rate and comment on one another’s responses. But we can go further. The results can become a social object – the aggregate data, or the insights and conclusions we build from it, can become open source – open to comment, modification, improvement by participants. And individual respondents can be social objects – befriended, trusted, or cajoled by their fellow participants.


It sounds pretty radical, right? Actually, the approach is a decade old at least. Its unheralded pioneers were sites like TheSpark.com, a personality test site hugely popular in the turn of the decade blogosphere. Users could fill out personality tests, assess themselves against the aggregate results, republish the results, pass the survey on, rate themselves against friends, create a profile, even create their own surveys.

Nobody called this “market research” of course. But people used it, and thanks to TheSpark and hundreds of other sites consumers now have certain expectations of surveys and survey data. They are used to being able to see the results and share them, at the very least. We as an industry came late to this party, still believing that our USP was that we gave consumers a voice. But they already had a voice, and interactivity is becoming a hygiene factor, not a selling point. If what makes us different as an experience is that we restrict participants, not empower them, that’s not a wonderful position for research to be in.

The problem is that interactivity isn’t great news for research practise. Allow respondents to interact and share and all sorts of problems arise. Representativity is critically threatened. Interaction effects skyrocket. Codes of conduct find themselves in a privacy nightmare. And incentives threaten to spiral.

These are all serious issues, and to me at least they suggest high-quality traditional research will always survive as a premium product. But let’s focus also on the positives of respondent interaction. It helps you understand the flows of influence and ideas in ways that would have been impossible before. It lets you harness the power of co-creation and the participant imagination. And it lets you turn your surveys into living objects by opening up your analyses and conclusions.


What does the world of interaction and social objects mean for engagement, though? Are social object surveys always more engaging? Not quite. What you’re actually doing is outsourcing engagement – relying on your parfticipants to engage one another. Which means that your study is only as engaging as its participants.

And not all participants are created equal. In any community – short-term or long – a pattern of participation inequality is likely to emerge: the notorious 90-9-1 rule which states that only 1% of community members will be heavy participants, 9% will be light participants, and 90% will simply observe from a distance, or lurk.

So immediately you’ve got problems at both ends of the pyramid. Your 90% are possibly unengaged – bored of the subject or too inhibited to contribute much. At best their engagement is passive. Your 1%, meanwhile, are probably over-engaged, exerting a distorting influence on the conversation via their passion for the project.

So what’s to be done? I’d propose four principles for creating engaging research which works as a social object: sociability, scalability, durability and playability.


Sociability is simply the decision to make elements of your project into social objects – enabling your research to be social in the first place. What layers of the work will participants get a chance to interact with? Each other? The responses? The results? Will you go all the way and turn your project into an open API, giving participants control of the structure not just the data?

Scalability is the principle of flattering the participation inequality curve – creating a project that’s appealing for the 90% as well as the 1%. In practise this will mean creating layers of appeal, but also structuring your rewards so they involve group goals not individual goals is a good plan.

Durability is the principle of maintaining engagement over time. If you are creating an ad or promotion to work as a social object, that might be a one-shot deal. But a research project is likely to have a duration – whether a week or a year, you need to make sure engagement isn’t front-loaded. Think in terms not of a discussion plan, but an engagement plan – perhaps using stories or games as an architecture. Where are the ‘Acts’ in the narrative you’re creating with participants? Where are the cliffhangers? What topics are the end of level bosses?

And speaking of games, Playability is the fourth principle – making your project into something that’s user-friendly, attractive, fun, and above all responsive to how the participant wants to use it. Because if you don’t provide fun, your participants will have fun at your expensive – ridiculing or gaming your projects. Market research, whether we like it or not, needs to become a branch of the entertainment industry – and social objects are the way we can get there.