Crossing the line with life event data | Twenty of Time

Crossing the line with life event data

Dystopian data practices for monetary gain

Friso van Dijk Friso van Dijk

What I remember most from my university graduation is the happiness of reaching a major milestone in life. Sharing it with my friends and family was an important part of that experience. It was posted to Facebook, I updated my LinkedIn profile, I got ready for my first job. What I didn't know was how marketeers would use that data to try and influence my life.

Life events come in many forms. Starting a family, buying your first house, graduating, they are all moments of large changes in our lives. What makes them important is not just the impact they have, but the changes we make because of them.

Humans are habitual creatures and changing habits is hard. Our resistance for change is lowest when a life event is going on. Our environment changes, leaving our lives open for new influences. In marketing, this is seen as an opportunity to bind a customer to their business. To insert themselves in our new habits.

It makes life event data very valuable. Knowing what someone is doing in their life is worth more than knowing their age or gender. Catching someone in the moment their lives are changing gives marketeers the opportunity to truly influence purchasing decisions.

There is nothing inherently wrong with this practice. As long as it happens with informed consent. But it's there that businesses have been failing again and again. They hide what they know about you, gather data in indirect ways and offer no reasonable way to opt out.

Unrelated data

In the most cited example of life event marketing, the New York Times discusses the American retailer Target. In 2012, they used statistical models to categorise customers based on their purchase data. One day, an angry man walked into a Target store with the message that his daughter got pregnancy-related ads in the mail. She was a teenager and in high school, he told them. She didn't need coupons for baby clothes.

Target followed up on the issue a few days later. The man explained he had a heart-to-heart with his daughter. He learned she was due in August, and he apologised for passing blame to them. But how exactly did Target find out about her pregnancy before her father did?

A Target employee illustrates this process, “Take a fictional Target shopper named Jenny Ward, who is 23, lives in Atlanta and in March bought cocoa-butter lotion, a purse large enough to double as a diaper bag, zinc and magnesium supplements and a bright blue rug. There’s, say, an 87 percent chance that she’s pregnant and that her delivery date is sometime in late August.”

Target looks for correlations in their data and uses them to classify customers. Customers who bought large purses may have a high probability of buying diapers a few months later. Find a lot of these probable relations and you can estimate if a customer is pregnant based on seemingly unrelated purchases.

They tested this form of customer targeting by sending out personalised folders. Pregnant women would receive coupons for the things they would need as an upcoming parent. They did not expect the backlash. Mothers-to-be were aggravated by how much Target knew about them. They hadn't told Target they were pregnant. It was creepy.

Instead of wondering whether they were invading their customer's privacy, Target decided to hide how much they knew. They started mixing the pregnancy advertisements with unrelated ones, such as lawnmowers. Now the customer could find the ones they were intended to without Target being creepy. The new approach turned out to be extremely effective. It was a huge success to Target's bottom line. They made a lot of money by hiding an invasion of privacy.

First period

Finding out about your daughter's pregnancy through ads may be due to a lack of information. But imagine something like this when your daughter turns 11. It happened in The Netherlands, where a girl's parents received an envelope from Always. It carried two menstrual pads and an information booklet about having your first period (Dutch). It's addressed to the parents, but it’s clear who the message is for.

It’s an example of how the data value chain works. The data source for Always traced back to the Royal Dutch Football Association (RDFA). Everyone in The Netherlands who’s registered at a football club is also registered at the RDFA. Even though the girl in question hadn’t been a member for years, her data was still being used somewhere.

The RDFA sold data with addresses and age ranges of its child members to data broker EDM. EDM then combined this with data from different sources and sold it to Procter & Gamble, the parent company of Always. By the time it was used to send targeted letters to girls around the age of their first period, nobody had stopped for a moment to wonder if they were doing the right thing.

When this hit the news, all parties repented and promised improvement. While I do not know about Always and EDM, recent reports show no betterment by the RDFA. They made headlines again in 2018 for selling bulk contact data of minors, with the clear requirement that buyers address it to the kid's parents. That's not improvement, that's an offense serious enough to trigger an investigation by the Dutch Data Protection Authority (Dutch).

Criminal opt-out

To complete our dystopian tale we need to discuss informed consent. It stands at the basis of how the law dictates acceptable use of our personal data by commercial parties. The concern exists that it's near impossible to not give your consent while still enjoying modern society. Janet Verseti, a Princeton sociology professor, decided to experience it for herself.

When she found out about her unborn child, she chose to hide her pregnancy from trackers. She chose to keep the big data machine out of her personal life. This is difficult when it comes to life events. Like the pregnancy in the Target case showed us, even random dots can be connected in unforeseen ways.

The first step for Verseti was to ensure there were no signs of her pregnancy on social media. If Facebook registers your life event, you will be tagged to receive ads relating to it. It’s one of the biggest assets to marketeers. With how embedded Facebook is in our society, it meant making some hard choices in how she shared her life with the people she cared about.

Next up was the bigger problem of not getting identified through her buying behaviour. Physical stores were easy – they accept cash. Online shopping proved to be a bigger challenge. To avoid being tracked by Amazon, Verseti used a privately hosted e-mail address and had everything delivered to a rented pick-up box. She and her husband paid exclusively with gift cards bought with cash.

To avoid online tracking of her behaviour, Verseti used the TOR browser. It's a browser designed to give the highest guarantee of anonymity. It hides your real location and comes with a lot of privacy benefits, but it gives up a lot of conveniences and, most importantly, speed. When a page finally loads in the TOR browser, you continuously have to click images of street signs and crosswalks for Google's CAPTCHAs. It's a slow and inconvenient way to use the internet and quite the trade-off to make.

When it was all over, she shared her experiences with the world. Her most astounding observation was that choosing privacy made her feel like she was doing something wrong. Buying a large amount of gift cards with cash could lead to investigations for money laundering. Having to be untrackable for physical mail delivery and using a new e-mail address for online shopping were not normal life experiences. Verseti said it best herself, “Those kinds of activities, when you take them in the aggregate [...] are exactly the kinds of things that tag you as likely engaging in criminal activity, as opposed to just having a baby”

Final thoughts

I started off saying there is nothing inherently wrong with using life event data, as long as there is informed consent. These three cases show that consent wasn't even in the picture. It was assumed by the companies, because why wouldn't someone allow the sale of personal data for someone else's bottom line? The sneaky collection and obfuscated use of this data is an ongoing invasion of privacy.

The current system is broken, not focused on the human behind the data but on the pursuit of wealth. As one insurance broker says it, customer segmentation "is not about being nice to customers. It's about making money."

What can we do against this status quo?