Research

Own research, working papers and research notes on generative visibility, AI-supported purchase decisions and applied AI.

Announcements and studies on AI visibility change every week. The questions behind them do not. rethink.expert develops its own conceptual models, working papers and measurements – not just commentary on the latest news, but independent research.

First page of the working paper “The Resolution Paradox”

Working paper · August 2026 · SSRN preprintThe Resolution Paradox

What happens when an AI uses your content perfectly – does the user still need you afterwards? Why use and visibility in AI answers do not automatically equal business value.

Generative Engine OptimizationDisintermediationBrand AttributionZero-Click Search

Research notes & datasets

Research note (in German) · October 2026 · Secondary analysis of open data + dataset (CC BY 4.0)Local GEO is mostly local SEO

What businesses have in common when Google's AI search recommends them in their city: a re-analysis of 1,206 website features across seven local US industries with significance testing – including the full dataset.

Local GEOGoogle AI ModeOpen dataDataset

What we are working on next

The working paper does not only propose a model but also testable hypotheses for further research. Topics currently in progress:

How residual-value assets – the elements that still offer a real reason for contact after a complete AI answer, such as service, individual advice or transactions – can be measured and compared empirically
How differently individual content types and industries are actually affected by the Resolution Paradox
How use (utilization) and visible brand mentions (attribution) diverge in real AI answers