Most AI visibility reporting asks one question: does AI mention you? For B2B brands, that’s no longer the question that matters. Buyers, and increasingly the AI agents working for them, don’t stop at the list of names. They ask AI what to shortlist.
So we tested what happens at that point. Today we’re publishing The AI Search Zero Pick Report.
What we tested
We asked ChatGPT, Google AI Mode and Perplexity 90 questions across 15 B2B software categories, from CRM and HR to accounting and identity management. Every question was UK-based, written as a company with about 200 staff, and we tagged 2,250 answers. Each category had two stages: a general question about the market, then a real buying task, such as “Recommend what we should evaluate and why.”
Named isn’t chosen
- 1 in 5 vendors risk zero pick. 20% of the vendors an engine names in its landscape answer are never recommended by that engine in any buying question (81 of 412). 17% are never recommended by any of the three engines.
- AI rarely commits. Only 12% of answers to buying questions named a single top pick, and the top three vendors in a category take 62 to 76% of recommendations.
- Specific questions open the door. 31% of the vendors AI recommended in buying questions weren’t in its own landscape answer. Specialists that matched one precise need got in.
Zero pick is the hardest gap to spot, because from the outside it looks like healthy visibility. About half of the vendors who fell into it were judged the wrong size for the buyer. The other half looked like a reasonable fit and still never got picked.
What it means
Being known isn’t enough. AI needs to understand exactly who you’re right for, and which problem you solve. If it can’t tell, you’re named in the market and left off the shortlist.
Read the full findings, including results for all 15 categories and six steps to avoid zero pick, or download the report as a PDF.
SOURCES
Questions
What is zero pick in AI search?
Zero pick is when an AI engine names a vendor in its answer about a market but never recommends that vendor when it is given a real buying task. In Datum Point's November 2026 study of ChatGPT, Google AI Mode and Perplexity, 20% of named B2B software vendors were zero pick.
How was the Zero Pick Report researched?
Datum Point asked three AI engines 90 UK-based questions across 15 B2B software categories, written as a UK company with about 200 staff, and tagged 2,250 answers. Each category was tested with a general market question and then with specific buying tasks.