3 ms·
I wonder (and really have no proof of this one way or another) whether this actually is successful a good bit of the time. HN readers are generally pretty thoug
by everdrive 6y ago
I wonder (and really have no proof of this one way or another) whether this actually is successful a good bit of the time. HN readers are generally pretty thoughtful with their purchases, but maybe a lot of Amazon customers really will buy two vacuums? Maybe they see there’s a nicer one, and just can’t help themselves?
- imglorp 6y agoFrom these examples we should conclude it works sometimes for emotional or artistic items subject to impulse buys, because Amazon sells billions and knows their business, but maybe not for utility items. The problem is, it's hard to distinguish for an algorithm. Vacuums no, but tools or kitchen gadgets, maybe
- hiq 6y agoExactly. Moreover, the ad about the second vacuum cleaner only has to be more effective than whatever random ad distribution was used previously, which would be a rather low bar.