Facial recognition is a pervasive and powerful tool, offering brands near-limitless possibilities when it comes to engaging their target audiences. But despite the benefits it can afford both brand and customer, is it just too, well, sinister? Dave Waller talks to futurist Mark Stevenson and investigates…
Marvin: I've been communicating with the ship.
Ford: What did it say?
Marvin: It hates me.
Poor Marvin. The manically depressed robot from Douglas Adams’ Hitchhiker’s Guide to the Galaxy series has one major burden: his brain. It’s the size of a planet, which has left him so bored that he now views anything that existence has to offer in the worst possible light. But while he may be paranoid, he can also intuit human emotions very well: everyone he encounters does end up hating him.
Here on Earth, computers are getting far cleverer too, but here they’re becoming an intimate part of our lives – they allow us to order anything on the planet to arrive at our house within a day or two, chat to friends and family around the world while streaming films at home, and even bag our shopping without talking to an actual human. There is, however, one technology that seems destined never to win a place in our hearts: facial recognition.
Like Marvin, it sees too much. And no one really likes it.
It can see it in your eyes
Facial recognition is an AI-powered biometric application that analyzes faces in photographs and compares the features it sees against an existing database. And it’s now part of our everyday: Facebook uses it so you can identify friends in photos, the iPhone X so you can unlock your phone without having to trouble your poor thumb to swipe.
But as is so common these days, the tech comes with a stark trade-off – between personal convenience and growing concern over broader rights and freedoms. And the risks are particularly stark. As well as the familiar question of how Facebook et al. are using your data, facial recognition takes us squarely into the realm of Big Brother: police in South Wales recently came under fire for tracking people at peaceful protests and then while out shopping, while Green Party peer Jenny Jones has written to the Met Police describing facial recognition as a ‘dangerously authoritarian technology’. In China, police now wear facial recognition sunglasses to curb protests of Parliament and blacklist people from events.
To most people who stop and think about it, this encroaching surveillance culture is Orwellian, dystopian, and definitely not ok, computer.
The reality is, however, that the technology is here to stay. Its appearance on the iPhone has already thrust it into legitimacy: as people get used to using facial recognition on their phone, other brands will surely jump on board with thousands of other applications. And the more people use it, the more of a research base the AI has to draw on, the more powerful and accurate it becomes.
Which leaves marketers in an interesting position. Here’s a technology that can surely be used to improve results. If you’re able to see who’s looking at your branding, you may be able to recreate the strengths of online marketing – of tailoring, targeting and measuring the impact of messaging – in the physical world, which could provide a much-needed boost to our beleaguered high streets, for example. Rival marketers will surely be jumping on board and profiting as a result. But the question is: Just because a technology is out there, does that mean you should get involved?
Facing the future
Mark Stevenson describes himself as a “reluctant futurist”. A former marketing director, he has since built a career looking forwards, advising global organizations – like Sir Richard Branson’s Virgin Earth Challenge – on “the big interplay of demographics, climate change and technology”. He describes his approach to consultancy as “insultancy” – often it involves examining his client’s strategy and then highlighting how foolish it is, given the troubling direction in which the world is moving.
“My general view of all technologies is that they’re not the answer, they’re the question,” he says. “The question is whether a given technology is going to help make the world more equitable, more sustainable, humane and just. I’m interested in how facial recognition technology could be used to do that. If you walked into a pub talking about it as a means of selling you stuff, no one would say it sounds like a good idea.”
Recognizing an opportunity
Making the world “more equitable, more sustainable, humane and just” are laudable aims, but you don’t need a brain the size of Marvin’s to appreciate that this is quite a big task. No one in facial recognition-based marketing is quite there yet. Instead we have things like TasteFace, an app that used facial recognition to determine how much a person loved or hated Marmite.
That does at least seem harmless enough. And other marketers have been experimenting too. Unilever is currently trialling in-store facial recognition technology that measures shoppers’ engagement with on-shelf displays of its Knorr brand. By using discreet cameras on shelves, it can analyze facial expressions to see and log their reactions to products, allowing it to work out which displays are appealing to which demographics. It’s the sort of measurable decision being made online all the time, only now in the physical space. And true to Unilever’s general bent, shoppers’ images weren’t recorded, and their privacy was preserved.
Elsewhere, GumGum, a ‘computer vision company’, worked with P&G to recognize the athletes used in the brand’s Thank You Mum Olympics campaign, allowing the brand to track the reach of the campaign, and to position an ad next to each picture.
Moving up one level of weirdness, Italian company Almax has created the EyeSee mannequin, which tracks the age, sex and race of shoppers and helps tailor the service to suit them. Meanwhile Clear Channel has been experimenting to determine if adverts can be altered according to an onlooker’s mood. One outdoor campaign used facial recognition to personalize car ads – the hoarding’s camera scanning number plates to identify passing car marques. If it spotted a Mercedes, it would inform its driver of the wonders of the BMW X5.
Turning our noses up
So is this just the warm-up? Are we about to see an influx of facial recognition tech, with audio ads on tube escalators blaring out ‘Stairway to Heaven’ on our ascent, because it knows we streamed Led Zep IV last night? According to our reluctant futurist, not quite.
“My take is that we’ll find it all too creepy,” says Stevenson. “Imagine leaving a car showroom, and the hoarding down the road suddenly advertises the car you’ve just seen – I think consumers will say that’s a massive no-no, that the showroom has gone too far, and it’ll prompt a massive backlash. Every brand wants their customers’ loyalty; they don’t want them to feel violated. We already have a massive issue of trust with companies, and this crosses that line.”
That line is, in part, about not wanting to live in a world that’s been manufactured to be more annoying than it needs to be. Then there’s the issue of privacy.
According to a 2015 survey of 1,000 UK shoppers, conducted by retail personalization company RichRelevance, people find the sense of being watched by technology ‘creepy’.
Nearly three-quarters (73%) of respondents said it was weird to enter a store and be greeted by name by a sales assistant they’d never met. Just over three-quarters (77%) said it was creepy discovering a salesperson knew your spending habits. And 68% said the same of targeted messages triggered by facial recognition.
This may change as the technology becomes more familiar, of course, but will we ever get comfortable with the feeling that our private data is serving somebody else?
With the onset of GDPR, people are slowly waking up to the fact that their behavior online is being tracked by cookies. We’re now served clear warnings about them on every new site we visit. Often that just means clicking ‘OK’, and carrying on regardless, because the concept of cookies is still too abstract. But would that be the same in a shop, where you know that if you walk in, all manner of surveillance, analysis and data gathering is about to unfold? After all, it’s impossible to install an ad blocker for your own face (for now, at least).
“The potential for brand damage here is enormous,” says Stevenson. “If people found out Apple was using the camera in their Air and then selling their face to various other websites, I think people would be like, ‘It’s over, you’re dead.’ It’d be the same if you found a store was using facial technology and selling that information to the people who own the banner ads on the side of bus stops. Or if Sainsbury’s was taking your picture at a checkout and using it to market to someone else. It’d be a massive PR disaster. That’d be it, the end of that brand. One exposé on Panorama or Newsnight and that’s it.”
So what of Stevenson’s question, of whether facial recognition can make the world a better place? It seems, in small ways, that it can. In 2015, Listerine ran a campaign using a facial recognition app that helped blind people ‘see’ smiles – their phone would vibrate when it recognized one. Users reported being very touched by the effect. Meanwhile domestic abuse charity Women’s Aid ran an outdoor campaign in which a model’s bruises disappeared from her face as more people watched the screen – a powerful way of showing that such problems can be tackled if we pay attention.
There are, of course, massive legal issues around all this. Facial recognition technology is as yet unregulated, but questions are being asked, and Parliament is currently debating the expansion of oversight of facial recognition and related data. The issue is, of course, that the technology is moving far faster than the legislation. So for any marketer concerned about personal freedoms as well as profits, legal guidelines will be of less use here than moral ones.
“There are clearly some applications where the technology can be used to make the world a better place,” says Stevenson. “The problem is that mostly it’ll be used for selling stuff: ‘Oh you look sad, here’s a chocolate bar.’”
It’s a depressingly Marvin-like view, but that doesn’t mean it’s not true. So is Stevenson right to be paranoid of the androids? We’ll find out soon enough.
For more on Mark Stevenson, follow him on Twitter.