Privacy, technology and advertising
Professor Dilip Soman discusses the use of facial recognition software
Recently a London ad campaign on an Oxford Street digital billboard raised some eyebrows. The billboard was equipped with facial-recognition technology that can tell with 90-per-cent certainty the gender of someone standing in front of it. This “skill” allows it to display a full 40-second ad for Plan UK, an international charity that helps children in the developing world, to women only. Men aren’t able to see the ad and instead are directed to Plan’s website. U of T News spoke to Dilip Soman, a professor of marketing at the Rotman School of Management, about whether this type of targeted data collection and marketing is an invasion of privacy.
Is this type of data collection an invasion of privacy? Is there anything an individual consumer can do to stop it?
I actually love the idea. I see it as an evolution of marketing -- something that is inevitable and something that could truly add value to consumers. The question about whether it’s an invasion of privacy is a separate question from the technology; it has to do with how the technology is used.
Let's think about the privacy issue more deeply for a moment. Guessing who you are and delivering an appropriate message is not an invasion of privacy. In the olden days, advertisers used different versions of ads in Vogue, Time and Sports Illustrated. That's because they had data that predicted the profiles of readers, and hence, they customized messages and content to match that profile. The next stage in the evolution is online; online marketers can get a good sense of what your needs are by your browsing behaviour and can deliver appropriate messages. I see the e-billboards as a natural progression of this basic idea of targeting messages to the right segment.
A violation of privacy occurs if your data are used in an intrusive manner, share it with others or somehow broadcast it. One thing I would worry about is the manner in which the e-billboard delivers messages. If the billboard delivers messages that could be potentially embarrassing in a public place, that might be an issue (but again, not a privacy issue per se).
A privacy violation would occur if the billboard used data that was not publicly visible. Let's say the billboard used face recognition technology to pull out your data from a grocery store loyalty program and delivered an ad for a hygiene product based on your usage patterns in the past; THAT would be a violation of privacy. Otherwise it is simply using data that is visible to anyone around the person.
Is this kind of targeted marketing the wave of the future?
It is indeed. The trick is for a) marketers to develop technology that does this in a cost-effective manner and b) to deliver messages those do not violate privacy and cause embarrassment. If these conditions are met, it is clearly a win-win situation. Marketers reduce wastage by focusing their advertising dollars on relevant audiences, and consumers do not get to see ads that they will likely never want to view in the first place.
One simple "fix" to the possibility of upsetting consumers is to allow them to opt-out. Suppose the billboard showed a standard ad, but had a button which could let the consumer control whether they wanted customized ads. This concept would work for online targeting, billboards, kiosks, cellphone advertising etc., and would go a long way in alleviating concerns about the potential “creepiness” of such targeted advertising.
Can this be something Canadians would soon see here as well?
That's a harder question to predict. I can certainly see why this would have appeal universally. I also think the younger generation in Canada is technologically savvy enough to want targeted marketing like this. The question is whether our markets are large enough to make the technology cost effective.
What are Canada’s rules around marketing procedures when it comes to marketing to the public? What generally do our laws say on what’s permissible and what isn’t?
As far as I know, this is permissible in our laws because- again, it’s simply a contemporary way of delivering advertising using principles that are age old. If this is illegal, then personally referring to women as "Madam" and men as "Sir" based simply upon seeing them and assessing their gender should also be illegal.
What is the potential for discrimination in this kind of marketing via this technology?
This is simply a technology; it enables whatever a marketer wants to do. So if someone wants to discriminate, yes they can do it. But they can also do it today -- so, if the question is "does the technology INCREASE the potential for discrimination," I'd say no. By taking control away from a human being, I would almost say that it reduces the randomness of certain individuals who might have discriminatory tendencies.