SEO Recap covering July 11, 2026: New research says AI visibility rankings are mostly noise

Daily summary of what matters in SEO, GEO, AEO, and AI search generated with Claude Code (beware of hallucinations)

---

Generative Engine Optimization (GEO) & Answer Engine Optimization (AEO)

AI Visibility Rankings Are Mostly Statistical Noise, New Research Shows
Source: Search Engine Journal

• A forthcoming IQRush preprint by co-founder Ron Sielinski argues that AI citation shares fluctuate because generative models add randomness, so dashboard rankings are snapshots of a moving target rather than fixed facts.
• Across 30 platform-topic tests on 3 engines, rankings needed 33-94 cited answers to stabilize. 3 tests (all on SearchGPT) never separated their top sites even after 125 questions.
• A prior study found SearchGPT cited Tom's Guide ~9.5% and Runner's World ~6.0% for running gear, but the 3.5 point gap fell within the margin of error, so one sample could not confirm a winner.
• Sielinski recommends measuring before and after content changes more than once, since a 3 point gain can fall within natural run-to-run variability, and Rand Fishkin advises vetting whether a tracker "shows their math."
• Engines differ. Gemini piles citations onto the same sites so each carries less independent information, while SearchGPT spreads fewer citations further. Top-10 sites still showed a typical margin of error of ~5 positions.