|
|
|
|
|
by thaumasiotes
764 days ago
|
|
> Demographics are used pretty heavily in market research, etc. This could be used both in targeting ads to specific people, and also in providing aggregate data to business and marketing partners I'm thinking about the workflow here. First, Discord profiles you based on your behavior. They conclude you are a 20-year-old male. Second, they show you ads appropriate for a 20-year-old male. Third, these ads do better than average because they match your behavior. We already have behaviorally-targeted ads. We've had them forever. How is introducing a level of indirection, where we infer age and sex from behavior and then decide on appropriate ads based on the age/sex construct, supposed to improve over deciding on the ads based on behavior? |
|