By Brian Tarran.

Online behavioural targeting has riled privacy advocates and lawmakers in the US – but any moves to restrict online tracking could have damaging consequences for research. Brian Tarran investigates. (Published in Research Magazine.)

Howard Fienberg has a busy couple of years ahead. As director of government affairs for the Council for Marketing and Opinion Research (CMOR), he mans the frontline in the fight to protect the US industry from legislative threats. Largely his work goes unnoticed – but that's as it should be. If agencies aren't staring down the barrel of a scattergun piece of regulation likely to blow a hole in the side of the good ship Market Research, you know he's doing his job well.

All of which is to say that just because the wider industry doesn't see these threats before they are extinguished, it doesn't mean they aren't there. Take the online privacy row currently raging in the US and, to a less vitriolic extent, in Europe. This is centred largely on behavioural targeting, a practice by which media publishers, ad networks and technology companies keep tabs on the websites people visit, the content they read and the ads they click on in an effort to deliver more relevant marketing materials to specific web users.

Is the outcry over this practice something that should concern researchers? The answer from the majority of those we spoke to could be summed up as 'No, why should it?' For Fienberg, though, the answer is a definite yes. Online privacy and the behavioural targeting debate will be a key issue occupying his time over the next two years or so.

"For the rest of this year Congress is out – they may come back for a lame duck session but they're not going to do anything," says Fienberg. "Next year though, at least on the House side in the Energy and Commerce Committee, Congressman Ed Markey is a big privacy zealot and he wants to develop and pass some comprehensive online privacy law revolving around behavioural targeting."

And it's here that the problems for research will likely begin, says Fienberg. "Because they take such a very broad view of what behavioural targeting is they don't define it very well, which ultimately means anything as simple as a cookie could be considered a form of behavioural tracking and something you need prior consent for."

Target: research
If it wasn't their job to get these sorts of things right, one could almost forgive lawmakers for having trouble drawing a clear distinction between behavioural targeting and research.

A bill before the New York State Assembly earlier this year defined "online preference marketing" – behavioural targeting in other words – as "the collection of data, including non-personally identifiable data, across websites or pages in order to define or predict consumer characteristics or preferences for the purpose of advertising delivery".

"If you listen to those words, that is basically the definition of advertising research," said Randall Rothenberg, chief executive of the Internet Advertising Bureau. True. But likewise they are an accurate – if basic – description of how behavioural targeting works.

So what separates the two? In a submission to the Federal Trade Commission (FTC) in April this year Fienberg noted that while both research and behavioural targeting companies track online behaviour in one form or another, research does so for the purposes of "aggregating groups and segments of the online population, not for targeting specific individuals".

It's this personal dimension to behavioural targeting that has enraged privacy advocates. Mark Cooper, the director of research at the Consumer Federation of America, explained: "We're interested in people who create behavioural records or profiles to target marketing to the individuals whose records and profiles they have. A research firm that tracks trends is not providing that sort of information to an advertiser or distributor."

But Cooper was speaking last November, and attitudes have shifted since then. Fienberg's FTC filing in April caught the attention of Jeff Chester, one of behavioural targeting's most vocal critics. Chester, who heads the Center for Digital Democracy, was none too keen on Fienberg's request that survey, opinion and marketing research be excluded from any definition of behavioural advertising, saying: "Researchers can't be given a free pass to push the limits of data mining in the digital era."

Taking Chester at his word, research is now on the hit list too – and worryingly he and his cohorts seem to have won legislators round to their way of thinking. On the upside, the FTC has already stated its preference for self-regulation when it comes to behavioural tracking, but Fienberg says it won't be long before senators and congressman start "turning the screws".

"Over the next two years you can expect the FTC to be under a constant barrage of pressure," he says. "Frankly, I think that in the new administration you are most likely to see the FTC try to coerce firms to develop practices that move in the direction of prior consent for any kind of tracking and that is going to be problematic for a lot of research. They are not going to require it, but they are going to use means to effectively coerce that out of the profession."

iSpy danger
We're entering the realms of supposition here, and its hard to know exactly what the industry will have to deal with at some point down the road. But it is likely that any proposed requirement for prior consent on tracking will call for a much clearer demonstration of 'consent' than websites currently ask of their users.

And as it stands, they don't ask much. Typically websites or internet service providers (ISPs) will explain in their privacy policies that certain information relating to online behaviours will be collected and in some cases passed on to third parties, though this will be in an aggregated, non-personally identifiable form. Users don't get the option to opt out of this – the only way for them to avoid it is to vote with their web browsers or wallets and not use those sites and service providers whose policies they disagree with. If they want to use that particular site or service, they are left with little option but to consent.

That's assuming of course that people even bother to read a website's privacy policy before logging on. Personal experience would say that most don't, while the popularity of social networks provides further anecdotal evidence that people are largely unconcerned about posting up the most intimate details about their lives for all to see. Ask them whether they would support tougher privacy laws though, and you're likely to get a different answer, says CMOR's Fienberg: "If you give them any kind of opportunity, folks are going to say, 'Well yeah, of course you should protect my privacy'."

But at what cost? Tracking online behaviour is fundamental to keeping the internet economy ticking – without the sort of data it provides, advertisers wouldn't have a clue where to target their spend so would pull their budgets, no doubt taking many popular websites out with them.

A losing game
Bill Tancer, the global research manager for online competitive intelligence firm Hitwise, says consumers often fail to see that what's good for advertisers can also be good for consumers. His company tracks search trends and reports on the popularity of websites by analysing the online activities of 25 million web users, and its this sort of data that enables companies to be "much more effective in their marketing", says Tancer. "If they're more effective in their marketing, it lowers the cost of marketing their products – and if that's true then it lowers the cost of products to consumers.

"It also enables companies to be more specific in terms of tailoring their marketing programmes, which means consumers shouldn't receive a barrage of advertisements that are irrelevant to them."

Hitwise data – like much of the online behavioural data researchers employ – is anonymous and aggregated, making the risk to personal privacy minimal. But the risk to the internet economy if deprived of this data is much greater. Legislators will have to weigh these risks against each other when they sit down over the next two years to decide what to do about online behavioural tracking.

But consumers must also be given the option to decide how much privacy they are willing to give up in exchange for keeping the internet ad-supported but largely free to use. If they come down on the side of the privacy advocates, so be it. In the long run it'll be the web users themselves that suffer the most, says CMOR's Fienberg. "Its going to take them a while to realise that what they've lost along the way is the efficiency that they get out of websites and their capacity to keep track of things for themselves – the benefits they get from the sites, not just the benefits those sites get from them."