1 free read left. Create a free account for unlimited reading + AI research.
Create free accountArticle
Mobile Dev Memo
This new approach simplifies campaign management for advertisers. However, it removes transparency regarding targeting logic and placement data.
Platform-driven algorithmic optimization is increasingly displacing advertiser-controlled targeting in digital advertising, and that shift quietly erodes advertisers' leverage. Eric Seufert's 2017 analysis builds on his earlier discussion of Google's consolidation of its ad channels into the UAC interface and Facebook's event-based campaigns. Both products act as black boxes that handle placement and performance optimization, which benefits advertisers lacking sophisticated measurement infrastructure, particularly in mobile, where attribution requires a fragmented set of third-party tools and specialist expertise.
The central argument is that the lack of transparency, though commonly criticized, is the lesser concern. Campaigns are either profitable or not, and the parameters produced by optimization algorithms would be unintelligible to humans without broader context. The more serious problem is data ownership. Facebook reportedly segments users by more than 100,000 targeting dimensions, and Google and Amazon hold comparably proprietary behavioral data. These dimensions are unavailable to other platforms, so each dollar spent against them strengthens that platform's targeting at the expense of performance elsewhere.
Seufert contrasts two scenarios. When advertisers supply targeting data, they learn from campaign results and can carry that knowledge across platforms. When the platform generates and retains the targeting intelligence, the platform improves while the advertiser learns nothing, creating a self-reinforcing cycle that makes the platform more attractive and other networks seem like starting from zero.
He references Antonio Garcia Martinez's Chaos Monkeys, which frames targeting and optimization as interchangeable along a preference curve. Seufert contends this holds only in a one-to-one relationship between advertiser and platform. For advertisers working across multiple channels, owning targeting data is more valuable than a platform's optimization ability because it preserves budget optionality and negotiating leverage. Advertisers who cannot see even which dimensions are used become fully dependent on a single channel.
Each row is a game or company this document names, each column a document, oldest first. A mark means that document names it. Select a row to see those documents.