DATA IS THE NEW BLACK
Algorithms, privacy and the future of shopping.
“When an e-commerce brand sends you an email that says ‘What’s new for you’,
the products really were selected for you.”
THE DAY Mark Zuckerberg answered questions before the United States congress, I decided to see what he had on me. Like millions of others, I wondered if my data had been harvested for some nefarious purpose. So I downloaded it all. what I got back — beyond party-planning group chats and friends of friends I’d quietly unfollowed — looked less like my psyche laid bare and more like the deepest, darkest regions of my credit card statements.
It was all there. Every online store I’d ever shopped from — and plenty I’d never heard of. Net-a-porter, Matchesfashion.com, Moda Operandi and Farfetch all had my contact information. They’d all reached out with advertising, and it was working — I was clicking on them. Psychographic targeting, remarketing (aka retargeting) and ‘lookalike audiences’ aren’t just the domain of shadowy political operatives. your friendly neighbourhood luxury brands are doing it too.
“I think the average consumer is highly aware of retargeting ... they notice when they visit a website and then see ads for that site over and over ...” says Rebecca Walker Reczek, a professor of marketing at Ohio State University. “what they may not be as aware of is when they get an ad for a website they haven’t necessarily visited in the past but that is being shown to them based on their pattern of behaviour online. this is the type of behavioural targeting that might slip under the consumer’s radar; they are being shown an ad based on an inference of who an algorithm thinks they are.”
One brand I’d been seeing a lot of was Tome, a New York label by Australians Ryan Lobo and Ramon Martin that makes vibrant workwear and dresses for the kind of sophisticated summer party you wish you’d get invited to more often. So I asked Nicki Wilson of CVR Marketing, who handles their digital ad buying strategy, why.the answer was simple. “if you follow Tome,” — which I do — “you’re going to get information about Tome in your feed. ”typically, I was seeing ads from Tome, not newsfeed posts. Targeting customers who already love your product, whether they’re existing shoppers, social media followers or those who’ve signed up for a newsletter is “really your low-hanging fruit”, wilson explained.
its users. that’s why my data showed I was connected to brands I hadn’t bought from but which I had browsed. Facebook also allows marketers to create ‘Lookalike Audiences’ of people who exhibit similar patterns of behaviour to their customers (Google has a feature called Similar Audiences; Pinterest has Actalike; Twitter has an algorithm with a similar function). So if you’re being served an ad for a label you’ve never heard of, it might be because their customers also tend to, say,‘ like’ Harper’s BAZAAR on Facebook, live in Australia and have an Instagram account.
These patterns aren’t spotted by humans; it’s all powered by machine learning algorithms, a form of artificial intelligence. For Amie Stepanovich, US policy manager at digital rights organisation Access Now, it’s what computers, not people, know about you that can be particularly uncomfortable. “We’re seeing companies know more about individuals than [the individuals] do about themselves, because they’re taking information that’s given to them, or that’s available in some other database, and using algorithms and other tools to make decisions about people based on it.
“I think that’s the next piece that we’re finally getting a look at ... [is] the manipulation,” Stepanovich continues. “People can be microtargeted through their preferences and manipulated … Knowing when somebody might be more likely to purchase something. Inferring how much money they might have to spend at a particular time ... It takes the process of purchasing and puts a lot of power in the hands of the retail market.”
While it might feel a little creepy to have a phalanx of formulas following your every move, I have noticed lately my digital shopping experiences are improving. wilson says her clients all use programmatic logic (automated decision-making based on algorithms) to build better, more relevant email marketing. when an e-commerce brand sends you an email that says “What’s new for you”, the products really were selected for you.
“The more data you have [as a marketer], the more in touch you are with your audience [and] the more sophisticated you can be in personalising that experience,” says Hana Abaza, head of marketing at Shopify Plus, which provides the e-commerce platform for thousands of startup and established brands. “i think more and more we’re moving in that direction.”
I actually enjoy being served ads for K.j acques sandals when I’m shopping for my new summer wardrobe. And I’m not alone. In a recent study, Walker Reczek found that “once consumers know an ad was targeted at them as a result of their click behaviour, they were more influenced by it … So that means consumers might feel more sophisticated after receiving an ad for a luxury brand ...A luxury brand could use behavioural targeting to make consumers feel more like the kind of person who buys luxury brands, therefore increasing purchase intentions.”
Yet the study found that only works if you know why you’re receiving an ad. If the ad doesn’t align with your interests, it won’t change your self-image. That’s why transparency is key, not just for consumers, but also for marketers. All this data is being used for the same thing any good sales assistant does the second you walk into a shop — profiling you based on your tastes. there’s one key difference, however: when you shop in store and a sales assistant is being too attentive or overly familiar, you can always walk out. Online, opting out is more complicated.
There’s a good chance you’ve been asked for your consent to be tracked by cookies quite a bit lately. we’re seeing privacy legislation such as Europe’s General Data Protection Regulation put power over how data is used back into consumers’ hands. Stepanovich thinks this is a great first step, but is wary that there are still opportunities for companies to misuse the massive troves of data collected.
Knowing big data can be used to manipulate me, I decided to do what everyone does on social media. I curated the companies that could speak to me, to create a more polished picture of myself. I combed through my Facebook preferences and deleted brands that reinforced habits I’d rather break. Candy Crush and Deliveroo went into the bin. But Tome, which serves me my dream summer party dresses? I let them stay connected. we’re all being profiled based on our profiles, so we may as well pick out a flattering angle.