Skip to main content
Content types earn very different citation rates per engine. Peec’s Feb 2026 analysis of over 1 million citations (full study) measured which content types reach the high-citation bucket (2.0+ for ChatGPT, 1.5+ for Google AI Mode and Perplexity), filtering out types under 5% of the dataset.
  • ChatGPT — listicles dominate. Listicles are just under one-fifth of all URLs and 52% of them are cited more than twice per answer. Product pages have high presence but weak citation performance.
  • Google AI Mode — no favorites, brand-owned content preferred. No content type reaches “high presence, high citations.” The top performer, category pages, sits just under 10% — roughly 5× lower than ChatGPT’s listicles. Among its most-cited pages, AI Mode leans to official brand sources: category and product pages (homepages also ranked highly but fell under the presence cutoff).
  • Perplexity — extreme selectivity. Listicles and articles reach the high-presence, high-citation zone, but at much lower rates than ChatGPT; only ~6% of all URLs reach the top bucket at all.
The takeaway is not “just do listicles” — it is that each engine rewards different formats, so benchmark per engine (see Citation rate benchmarks by engine). Note that GPT-5.6 later shifted ChatGPT retrieval toward product pages — see GPT-5.6 retrieves product pages over listicles. Why listicles fit AI answers structurally: Why listicles get cited. Source: What does a good citation rate look like? Benchmarks from over 1 million AI citations, Peec (Tom Wells), Feb 2026.