3 ms·

> In reality, people don't realize how often they buy the same and similar stuff again and again. Yeah... this is not true for the vast majority of non-consuma

by tomrod 4y ago

> In reality, people don't realize how often they buy the same and similar stuff again and again. Yeah... this is not true for the vast majority of non-consumable products without some sort of time delay. If you bought 20 Zebra gel pens (best pens on the market, no affiliation just satisfied consumer) you don't need 20 Zebra gel pens for a sale refraction period of at least an hour. Immediately re-advertising is a poor advertising model, conceptually, and I'd really need to see a lot of evidence that this is wrong across a wide variety of products to move away from that prior.

saurik 4y ago
And like, a great salesman wouldn't offer you more of the same product but try to either expand into similar products (you like gel pens? here are some mechanical pencils) or hit you up with complimentary products (what good are gel pens without a nice squishy writing mat? and i bet you're going to need a lot of paper!). Amazon sort of handles these things in the Amazon UI with "people who bought X also bought Y" but they really should be showing that for their ads: instead of "here are results for your recent search" (which likely shows you the thing you already got) it could do "here are things people who buy stuff near the top of your recent search also buy". Regardless, as far as I am concerned, the best commentary ever on this topic was CollegeHumor doing the Targetted Salesman skit, where the ad engine is represented by a human salesman shoving products at you all day: https://m.youtube.com/watch?v=KbKdKcGJ4tM https://m.youtube.com/watch?v=KbKdKcGJ4tM
FrenchDevRemote 4y ago
Well I guess statistics don't agree with you or Amazon wouldn't make that much money
- tomrod 4y ago
  Amazonians and other big techies would absolutely agree that the strategy space is undersearched, which puts your assertion's assumption at risk. Took several years to even consider the statistical power of ad effectiveness.[0] [0] https://www.jstor.org/stable/26372642 https://www.jstor.org/stable/26372642

by tomrod 4y ago

saurik 4y ago

And like, a great salesman wouldn't offer you more of the same product but try to either expand into similar products (you like gel pens? here are some mechanical pencils) or hit you up with complimentary products (what good are gel pens without a nice squishy writing mat? and i bet you're going to need a lot of paper!). Amazon sort of handles these things in the Amazon UI with "people who bought X also bought Y" but they really should be showing that for their ads: instead of "here are results for your recent search" (which likely shows you the thing you already got) it could do "here are things people who buy stuff near the top of your recent search also buy". Regardless, as far as I am concerned, the best commentary ever on this topic was CollegeHumor doing the Targetted Salesman skit, where the ad engine is represented by a human salesman shoving products at you all day: https://m.youtube.com/watch?v=KbKdKcGJ4tM https://m.youtube.com/watch?v=KbKdKcGJ4tM

FrenchDevRemote 4y ago

Well I guess statistics don't agree with you or Amazon wouldn't make that much money

tomrod 4y ago
Amazonians and other big techies would absolutely agree that the strategy space is undersearched, which puts your assertion's assumption at risk. Took several years to even consider the statistical power of ad effectiveness.[0] [0] https://www.jstor.org/stable/26372642 https://www.jstor.org/stable/26372642