Assistants often disagree on the leading brand
A pooled ranking can conceal substantial differences between assistants answering the same buying question.
- Published
- 2026-08-14
- Revised
- 2026-09-01
- Author
- Cleotic Research
- Method
- cleotic-visibility-method-v2
eligible questions with a split winner
Denominator: 136 eligible questions
average top-five overlap
Denominator: 90 eligible assistant-pair comparisons
assistants in the retained panel
Denominator: full-coverage assistants only
44.1% of eligible questions had no single brand leading across every retained assistant. The denominator is 136 questions with an eligible cross-assistant comparison.
Across 90 eligible assistant-pair comparisons, the average overlap between top-five brand lists was 74.4%. The pooled ranking is therefore a summary of different answer sets, not a result reproduced by each assistant.
Use the combined table as a market-level benchmark, then inspect assistant-specific rankings before deciding where a brand is consistently visible. These comparisons describe the fixed panel and question set; they do not predict future answers.