“London-based angry feminists over 30”: why we need to talk about the sexism of online ad profiling

You are a 29-year-old male? Buy these trainers! You are a 29-year-old woman? Download this dating app for “older women”!

NS

Sign Up

Get the New Statesman's Morning Call email.

This should be the year when sexist advertisements end. The Advertising Standards Authority will start cracking down on gender stereotyping as of June 2019. Fighting sexism in the public discourse is all very well, but in today’s world where advertisers are tracking us to target us with specific ads, who is thinking about the sexism of the advertising industry’s obsession with profiling us?

Think about those pesky online ads you get based on what you may assume is just your browsing habits. In fact, in order to target us, the advertising industry needs to profile us. And when it comes to profiling, gender is still the easiest way to put people into boxes.

So, the ads we see online reflect what is expected of our gender and perpetuate the very stereotypes that the ASA is trying to fight. You are a 29-year-old male? Buy these trainers! You are a 29-year-old woman? Download this dating app for “older women”! Unlike a broadcast advert, where the sexism has been depressingly overt for decades and has been called out by campaigners for just as long, with a narrowcast advert it is much harder for us to know when and how gender assumptions are behind the adverts we see. The content itself might not be controversial or sexist, but the reasons it was put in front of you, and not someone else, might be very sexist.

When profiling us, companies are not interested in our names or who we really are: they are interested in patterns of our behaviour that they believe match an audience type, like Daily Mail readers, divorced fathers of preteens with a high income, newlywed couples more than three months but less than a year, women between 25 and 35 living away from their family, single mums in Liverpool… Needless to say those audiences types carry their fair share of assumptions and prejudices because ultimately they help companies make quick decisions about us.

So to target us more efficiently, the advertisement industry relies on a very binary vision of the world. We see this quite well with Facebook: they offer their users dozens of different options to define their gender identity, but go to the “ad manager” section – the ones designed for businesses wanting to create targeted ads – and the gender diversity is gone. Businesses are only offered to target users based on “male,” “female” or “all.”

But the problem is that the consequences to this profiling goes beyond advertisement. We know for instance that Google is less likely to show ads for high-paid jobs to women than they are to men. This is not because someone at Google makes the deliberate decision to prevent women from looking at high-paid jobs. It is because Google – like an increasing number of companies that need to make instant decisions – relies on artificial intelligence. And if there is one thing artificial intelligence is good at it is reproducing the behaviour and prejudices of those who designed it and those who gathered the data – that is, white men. (And yes, that’s me making gender assumptions - not nice is it?) 

We need to be clear that a data driven world – where artificial intelligence makes decision based on simplistic profiles about us – isn’t going to solve prejudices: it’s going to perpetuate them. If you’re a woman, or trans, or non-binary, chances are that you’re either going to be “binarised” into simple male/female categories, or you’re going to be targeted based on even more assumptions about what shower gel a trans person should buy, or where people who identify as queer should go on their next holiday.  Either way, how the advertisement industry targets us all based on its knowledge/guesses/assumptions about our gender has to be the next frontier of tackling endemic sexism in advertising.

If we are getting rid of the billboards showing women happily ironing their husbands’ shirts, let’s not at the same time just replace it with a 21st century version of a 20th century problem.

Eva Blum-Dumontet is research officer at Privacy International.