[kevin_thibedeau]

Whenever this issue comes up I feel the need to relay a story about a meeting with a databroker in 1998 who was tracking menstrual cycles using purchasing records of a wide variety of consumer goods. They will track you to optimize their manipulative, targeted advertising whether you have an invasive app or not.

[throwway120385]

What's different is that the period tracker apps have women provide that information pretty directly. The data broker probably had to do a lot of work and then estimate with a confidence level. So there's a significant difference in level of effort there, and then because of the estimate the data becomes much less valuable because it's only plausible it's not certain.

This argument gets trotted out whenever people talk about privacy and protection of personal information. I hear it when people discuss LPR systems, cameras, and the like too.

These data-gathering applications of technology should be treated specially because they make it extremely easy to gather high-confidence data about individuals. I bet your data broker couldn't predict when someone was ovulating or whether they were pregnant or count how many miscarriages they had just based on purchase history? This is all data that period tracker apps capture pretty directly. The miscarriages one is particularly bad because that data could be misconstrued as abortion and in US states where that's illegal it could get you into real trouble. There's a small but vocal subset of the population that seems to want to punish women for anything other than a healthy delivery. But when my spouse and I were trying everyone we talked to had at least one if not more miscarriage stories.

[annzabelle]

I've heard rumors, maybe that far back, of women getting mail ads for baby clothes a couple weeks before having a positive pregnancy test.

[vasco]

You are misremembering, what it you're referring to is the Target pregnancy prediction score which was advertised a lot in the early 2010s as the way "big data" was going to change the world because they figured pregnant women bought similar things

https://www.forbes.com/sites/kashmirhill/2012/02/16/how-targ...

[m463]

I wonder about not only baby or feminine supplies, but also snack coupons and dating apps

[kevin_thibedeau]

Any behavioral pattern with a 28 day correlation will pop out. Once you get a rough estimate, you can aggregate subjects with dates shifted into alignment to find more subtle correlations.