Filled marks: named in at least 2 of 3 scheduled repeats. Open marks: below that threshold. Dashed marks: excluded answers leave the threshold uncertain. Counts and intervals follow for each question.
Answer capture date: 2026-09-22.
By engine
ChatGPT
9 of 30 valid answers, 30.0% (Wilson 95% interval 16.7% to 47.9%)
Claude
5 of 30 valid answers, 16.7% (Wilson 95% interval 7.3% to 33.6%)
Perplexity
12 of 30 valid answers, 40.0% (Wilson 95% interval 24.6% to 57.7%)
Google AI Overviews
4 of 30 valid answers, 13.3% (Wilson 95% interval 5.3% to 29.7%)
Google AI Mode
6 of 30 valid answers, 20.0% (Wilson 95% interval 9.5% to 37.3%)
By buyer question
Each count is over valid answers to the exact question. Excluded answers are outside that base.
What is the best CRM for a growing B2B sales team?
0 of 15 valid answers, 0.0% (Wilson 95% interval 0.0% to 20.4%)
Question 1. Excluded answers: 0. Type: best option.
Which CRM is best for managing a complex sales pipeline?
7 of 15 valid answers, 46.7% (Wilson 95% interval 24.8% to 69.9%)
Question 2. Excluded answers: 0. Type: specific workflow.
What are the top CRM tools for a mid-market company?
15 of 15 valid answers, 100.0% (Wilson 95% interval 79.6% to 100.0%)
Question 3. Excluded answers: 0. Type: top tools.
What are good alternatives to a traditional CRM?
0 of 15 valid answers, 0.0% (Wilson 95% interval 0.0% to 20.4%)
0 of 150 valid answers, 0.0% (Wilson 95% interval 0.0% to 2.5%)
0 of 150 valid answers, 0.0% (Wilson 95% interval 0.0% to 2.5%)
Release and method
Release
2026-09, version 2026-09.1
Published
2026-09-23
Answer capture date
2026-09-22
Source capture dates
2026-08-28
Engines
ChatGPT, Claude, Perplexity, Google AI Overviews, Google AI Mode
Configuration
10 fixed target questions, 5 engines, 3 repeats. Five controls were run outside the scored base.
Method version
v1.1
Excluded answers
0 target answers excluded; exact reasons are listed. Controls were run outside the scored base.
Commissioned engagement
No commissioned engagement is declared for this category.
Exact names and listed aliases are matched without case sensitivity at word boundaries. Plain word names require a listed qualifier nearby in the same sentence or the canonical domain among cited URLs. A vendor citation counts only when that answer also names the vendor under this rule.
Verbatim answer text is retained privately with the measurement input and import audit. No answer quotations are published here.