Do brand-name prompts measure discovery?
Supplying a brand name in the prompt tests how the answer discusses that brand; it cannot establish that the engine would discover the brand unaided. The distinction matters in the July 2026 Volume II study: 11 of 16 absences among eight highly named companies occurred on comparison-shaped questions. Being found in a category and being included in a named-rival comparison were different observations.
What we measured
Measured. Volume II captured ten questions per company, once per question, on one engine from 18 to 23 July 2026. The convenience and challenger sample covered 85 B2B software companies and produced 616 absence records. It was not a representative sample of all software brands.
Nine companies were named in seven or more of ten answers. Eight of those nine had any absences, contributing sixteen absence records; the remaining company had none. Eleven of the sixteen absences were classified as comparisons. The published rule used “vs”, “versus” or “comparison” and gave that shape precedence over other shapes. Eleven of sixteen describes this small absence group, rather than the share of all answers that mentioned a brand.
The Volume II data and classification method let a reader inspect the question shapes and their bases. Reasoned. A question that names a rival can examine a competitive decision after the buyer already knows the names. This study did not test a content intervention or show that adding comparison pages would improve a result.
Our answer-study method, v1.1.
What this means for you
Label branded and unbranded questions separately in your report. An answer to a supplied name can reveal factual descriptions, comparisons or omissions, but it is a different test from asking for suitable suppliers without giving their names.
Retain both kinds when they serve real buyer decisions. For discovery, ask the category and need. For consideration, preserve the actual rival or product pairing. Do not combine the two into an unlabeled discovery score.
Check the base before interpreting a comparison result. The sixteen absences here were selected from a small high-visibility group; they do not predict the absence mix for your company.
Also asked as
Do brand-name prompts measure discovery?
Related
Next step
Try Free 10-question check (one engine) with an unbranded buyer question if your aim is discovery, and label any separate branded question as such. No email is needed; runs are limited per network and per day, and by the site's daily allowance.
When a branded answer contains a specific factual mistake, Fix one answer, $490 can address one answer without turning its presence into a discovery claim.
Prepared by the Broadcastwell team.
Page owner: Priya Nair, Management Consultant. Corrections: hello@broadcastwell.com