3 ms·
I worked on ad load optimization and founded a company that specializes in ad load optimization, here's an explainer video: https://www.promoted.ai/videos/unifi
by andrewyates2020 4y ago
I worked on ad load optimization and founded a company that specializes in ad load optimization, here's an explainer video: https://www.promoted.ai/videos/unified-ads-search-and-feed-optimization https://www.promoted.ai/videos/unified-ads-search-and-feed-o...).
The cynical comments here align with my experience: when leadership demands now revenue, there is an iterative game where the only correct answer is "increase revenue now." If you run enough A/B tests and ask this question enough times over a large enough organization, more ads and 3p always result.
Most companies are trying hard to produce a great user experience. However, it's hard to measure subtle degradations to buyer experiences, especially when those degradations happen after the purchase or quality metrics corrupted by motivated sellers or advertisers. This is one reason obsession with A/B testing drives this poor user experience: it's hard to measure. Revenue now is easy to measure.
Another aspect you may not see as a buyer is that when the market is down, sellers are SCREAMING at their platforms to fix the problem. Same iterated game: give boost (discount, ad credits etc.), less screaming (for that team right now). What you see as bad as a buyer may be trying to appease sellers.
- lastpingstandin 4y agoI worked for Amazon on a team that touched this, and I can confirm this is basically true. For recommended products, Amazon's algorithm prefers widgets that generate more revenue (which sponsored products has a huge advantage with), and thus those products typically get ranked higher.