Surveillance: Google knows you’re pregnant… before you do!
On social networks, we regularly hear about it. Google would know that women are pregnant before they themselves know. We also remember this terrifying story of a dad who had complained that his teenage daughter was receiving targeted advertising around motherhood (diapers, lotions, etc.) In the end the young girl was indeed pregnant and the Target chain of stores – which developed “pregnancy prediction score” – so knew she was hearing a happy event even before her dad knew about it.
But this story, which dates from 2012, is not an isolated case. It’s even worse 11 years later since if this young girl had indeed searched for products or keywords related to pregnancy, the algorithms are now suspected of going even further…
Probability or algorithm?
Several women’s testimonies are similar. They suddenly receive advertisements on the Internet related to motherhood: strollers, baby food, clothes, accessories, etc. A few days or weeks later, they discover their pregnancy.
There are three possible explanations for this.
The first is probability. These advertisements are addressed to all women between the ages of 25 and 35. Age is very easy to get. The advertiser crosses his fingers hoping that Madame wants a child. In short, it’s haphazard luck and those who become pregnant between the time they receive advertisements and the time the test is positive are inevitably surprised. But it makes sense in the end.
The second explanation is more convoluted. Data scientists agree that pregnant women become very sensitive to smells. They don’t like “scented” items. Therefore, if a woman suddenly stops buying perfumed soap or perfumed lotions, it is very likely that she is pregnant. This behavior is hard to notice from the woman’s perspective, but it’s easy to detect for a store that knows everything you’re buying.
Google is not your friend…
There may even be less obvious patterns or a combination of patterns (different food, particular purchase, etc.) that would allow an algorithm to know whether a woman is pregnant or not. Finally, there is incidental Internet research. Madame is looking for things on Google that conclude with a symptom of pregnancy: nausea, cramps, constipation, cold feet.
She does not realize it immediately, but on her side Google cross-checks this data with the age of the person and bingo!