3 ms·
Here is a link to the actual report which the article is discussing: https://au.reset.tech/uploads/resettechaustralia_profiling-children-for-advertising-1.pdf h
by RileyJames 5y ago
Here is a link to the actual report which the article is discussing: https://au.reset.tech/uploads/resettechaustralia_profiling-children-for-advertising-1.pdf https://au.reset.tech/uploads/resettechaustralia_profiling-c...
In the report it states the ads were not run. I’m pretty anti-Facebook anything, but this feels like an attack aimed to generate anti-Facebook headlines off the back of their flawed methodology.
I find the idea that any company is going to ‘exploit’ the capability to purchase advertising in $5 batches to target 13 - 17 olds with products such as cigarettes, vapes and gambling a little far fetched. Particularly for any company licences to sell said products. Maybe middlemen / affiliates, but even then the cost associated with creating facebook advertising accounts for such small budgets doesn’t seem cost effective.
It seems like their methodology has been deliberately chosen ($5, non-branded, trademarkless adverts) to avoid automated filters by flying low.
If they did run, they’d likely to be flagged by human users, escalated for moderation and removed.
It comes off very calculated.
- michaelmrose 5y agoAdvertising is virtually always manipulation. Logically if you want to be ethical you just won't let them target minors at all or you will be incredibly circumspect in whose human reviewed ads are run.
- helsinkiandrew 5y agoI've always assumed there is some kind of human moderation in the review process to stop things like this happening, after that they are approved and can be shown on websites as soon as the advertiser wants. Relying on viewers to flag (presumably teenagers interested in smoking or gambling) doesn't sound like a good idea. This does just sound like the interests ("gambling", "cigarettes & vaping") should have age appropriate flags so they can't be combined with the specific age groups.