FEATURE1 March 2009

People who need people

Web 2.0 networks like Facebook and LinkedIn look like perfect providers of survey samples. But have they got the research expertise to carry it off? Richard Young investigates

?When he was asked why he robbed banks, notorious 1930s stick-up man Willy Sutton simply replied, “Because that’s where the money is.” Researchers are offering the same justification for an interest in social networks like MySpace, Facebook and LinkedIn. Why look there to find candidates for online surveys? ‘Because that’s where the people are.’

Social networks are the very embodiment of Web 2.0, and they’ve been phenomenally successful. Facebook, for example, is only just five years old but already has over 160 million users. LinkedIn, which targets business people and boasts director-level members from most blue-chip companies, has 35 million signed up. Even the most basic registration for these sites requires users to offer some demographic information. Those taking an active interest in their membership reveal massive amounts of data about themselves.

That means they’re a potential goldmine for anyone looking to round up an online panel. Not only could you interrogate the database to target research projects at tightly-defined, self-selecting groups – these users often come from traditionally hard-to-reach audiences such as the 18-to-34 year-olds, senior managers and IT decision-makers. Bingo. And for the social networks – it’s a great way to monetise those eyeballs.

False dawns
But while a marriage between a ready supply of potential respondents and a market research industry desperate for younger or more business-focused respondents seems made in heaven, there are problems.

It’s one thing to filter a large universe of users and let MR firms loose on the ones they really want – but quite another if privacy fears force those same users to rebel against the sale of their demographic data and resist commercial exploitation of their community. That’s before these sites even start to consider the regulatory hurdles of selling access in this way.

“If a social network wanted to get really serious about market research it would have to collect more data about its users – and that’s data users are reluctant to give,” says Michel Guidi, European MD for Survey Sampling International (SSI). “I’m also curious about the tolerance of social network users for material being pushed into their profile pages.”

That’s a lesson the more youth-oriented sites have already learned. In 2007 Facebook launched Beacon, a service that allowed partner organisations’ to track Facebook users’ activity on their web sites, allowing the social network to refine its ad targeting. Privacy advocates and users were united in their objections, and the system was shelved. Then, in December, the site launched Facebook Connect, where users can log on to partner organisations’ websites using their Facebook username. The social network is treading carefully around privacy issues this time – but even if it does get access to its members’ activity logs at third-party locations, it’s unlikely to do much more than target advertising more intelligently.

But there have been more research-oriented attempts to monetise members. Facebook Polls – a system allowing companies to target users with a single, multiple-choice question – was launched in 2007 but dropped early in 2009. It looks ready to make a comeback sometime around the end of this quarter, according to its UK head of trade marketing and market development, Trevor Johnson. Facebook founder Mark Zuckerberg proudly showed off the new version of the system at the World Economic Forum in Davos, although at the moment it remains predominantly an internal research tool.

LinkedIn runs an almost identical service allowing you to poll your own network free and targeted groups for a small fee. But these single-question offerings barely even scratch the surface of genuine research and could never replace online panels.

The real deal?
Much bolder claims about generating proper research samples have been made for the LinkedIn Research Network (LRN), “the World’s Premiere B2B Sample Provider for Online Surveys” (the quote – and caps – are theirs).

“Protecting users’ privacy – and not annoying them – is only part of the problem for social networks interested in making money from research”

Originally conceived as a investment tool – allowing analysts, for example, to question experts in a particular industry sector in the way the Gerson Lehrman Group does – the system is now advertised as a bona fide sampling system, complete with user opt-ins, screening and membership of the Council of American Survey Research Organisations.

Although it clearly sees a valuable niche in creating raw samples, LRN has so far partnered with established market research agencies (such as Phoenix Marketing International and OTX) to exploit the opportunity.

But it, too, is said to be moving extremely cautiously. “While they show good intentions, it remains to be seen how successful they’ll be at the actual market research side of the equation,” says Simon Chadwick, interim CEO at Peanut Labs and former CEO of NOP World. “The main asset they have is their users – but they’ll want to keep them happy. So it will be interesting to see what rules they put in place for research clients, like the length of surveys.”

Dan Shapero, director of business services for the LRN, told Media Post at the October launch of LinkedIn Surveys (part of LRN) that initially, at least, users would only get four questionnaires a year. “They will be offered rewards to share their behaviour like cash incentives, donations to a charity of their choice, and merchandise,” he said. “The other reward is that this research might be used to improve the industries they work in.”

Research spoke briefly to John Cotronis, LRN’s Head of Sales & Account Management, who explained that the company was only interested in genuine research projects – and was firmly rebuffing “amateur” or small-scale requests that might not justify users’ allocation of time.

But protecting users’ privacy – and not annoying them – is only part of the problem for social networks interested in making money from research. The fact is, they don’t really know much about it.

“They’re great companies and they could become pretty much what they want to over the next couple of years in terms of business models – and we think there’s a lot we could do together,” says Guidi. “It’s true that they have access to, and information about, some very hard-to-find respondents. But while that’s an important piece of the jigsaw, it’s not everything. You need the right tools, expertise in how samples work, good client interface – you need to understand the industry.”

In any case, research isn’t core to the social networks’ business model, and that’s likely to make it much harder for them to devote time and resources to becoming experts. “Web 2.0 means they’re all about openness and creating opportunities for people to interact on their platform,” says Guidi. “Where does administering and interpreting a 20-minute survey fit into that business model? It doesn’t.”

And he’d get no argument on that score from Trevor Johnson. “Our business model is still advertising and big brand deals,” he says. “Yes, we’ll do our own research to make sure we’re on top of our users’ needs and maintain the quality of the platform. And we have Facebook Polls and Lexicon, our buzz-tracking service. But in the short term, it’s safe to say we’re not in the sampling business – although this is a fast-moving industry, so who knows about what it’ll look like in six months.”

Where worlds collide
That’s not say there aren’t already massive opportunities for both sides in this equation. Why wouldn’t a research firm looking to assemble a high-level panel not talk to LinkedIn Surveys? Facebook’s Engagement Ads allow its clients to understand key demographic information about the people clicking their links – and since advertising can be tightly targeted on profile information provided by users, that creates a great opportunity for sampling companies to reach the right people.

“You need the right tools, expertise in how samples work, good client interface – you need to understand the industry”

It’s an opportunity the specialists have lost no time in exploiting. “We have already developed a Facebook application – it’s early days, but it looks very interesting,” says Guidi. “We weren’t sure about it at first for precisely the reasons we’re sceptical about social networks as sampling organisations – the fit in terms of the user experience isn’t obvious. But there’s certainly some potential.”

One company that claims to have cracked the social networks for research samples is Peanut Labs. It has addressed the privacy and intrusion issues by partnering with application developers and community organisers within the social networks. Respondents are incentivised to complete comprehensive questionnaires using the virtual currencies promoted by these applications; the developers are given a share of the revenue for assembling the panel.

By the end of this year, CEO Simon Chadwick, claims the company will have more than 200 of these sub-networks signed up to Peanut Labs – it currently boasts a universe of 37 million users; ten per cent of those have been fully signed up as panellists.

For the social networks interested in market research revenues, the decision is actually deceptively simple. Do they emulate Google? Or do they want to be Microsoft? Google’s model is to provide a free and open platform for users – in search, video, communities and productivity apps like email – and then offer highly targeted advertising based on users’ own preferences and environment. Microsoft, on the other hand, sells its operating system and applications.

Right now, Facebook’s Trevor Johnson is suggesting that his company is more like Google. “We have an open platform, which means anyone is free to set up groups or pages or write applications that help them undertake market research,” he says. “For example, an application or group could be advertised in a very targeted way to appropriate demographic groups to encourage them to join a research panel. Stripped down, it’s just a fantastically cost-effective way of recruiting your sample.”

LinkedIn – which coincidentally charges users for a premium service – looks far more like Microsoft. It’s bolting functionality right into the operating system and hoping to sell access in a much more structured way. Whether it succeeds or not will come down to the quality of the samples it provides the research industry, and at what cost. That remains a function of how engaged its users are – and whether it can crack issues that research specialists have been grappling with for some time.

“Our expertise comes in areas like the Respondent Preservation Initiative, which aims to ensure that more people want to take and complete surveys,” says Guidi. “We are thinking hard about all this stuff, all of the time [which the social networks can’t do]. But allying market research expertise to the possibilities presented by the social networking platforms could be very interesting indeed.” Interesting and mutually profitable? Only time will tell.