3 ms·
> Why, oh why, would a company built on ad revenue want to reduce ad spend? > These companies are perfectly positioned to make ads more efficient - and they hav
by xixixao 7y ago
> Why, oh why, would a company built on ad revenue want to reduce ad spend?
> These companies are perfectly positioned to make ads more efficient - and they have absolutely no reason to do so.
The question is not really asking about reducing ad spend, it's asking about reducing advertiser costs per outcome. It's pretty obvious that if you decrease costs per outcome, return on ad spend will rise, and advertisers will spend more. In fact, more advertisers will want to get in on the game (as costs decrease, ads become profitable for more and more advertisers), and this pushes costs up. This eventually creates an equilibrium, but even at that point, it is still the case that decreasing costs increases revenues for ad networks.
Attribution is a really, really hard problem. Ideally you want to do a Lift test, and ideally you want to do it across all your ad spend channels. To do this, you still need the attribution mechanisms ("tracking" in the article).
So I don't find the article coherent. It points out deficiencies in the system (which there are many), it falsely claims that big players have no incentive to improve, and it makes it sounds like all these advertisers are just burning their money. It makes it sound like pervasive tracking is a problem, which it might be for other reasons (privacy, unintended consequences), but certainly is not making ad tech worse.
- tomlockwood 7y agoAuthor here. I think your response only holds true if both the market acts rationally (which isn't the case if people are optimizing-for-resume), and the benefit from claiming more conversions isn't as low hanging as other "efficiency" gains. The "attribution mechanisms" you mention are precisely one of the drivers for the pervasive data collection that is in AdTech.