Reply in "Fictional Kalvenor: choose three checks"

If I could run only three checks, I would choose coverage, evidence, and stability.

First, I would run a small prompt map across the most important buyer intents in Kalvenor's category. That tells me whether the omission is isolated to one wording or whether the brand is absent across the category language buyers actually use.

Second, I would inspect the entity and source graph around Kalvenor: its own pages, indexability, product and category statements, and the independent sources that describe or compare it. That tells me whether an engine has clear, corroborated material from which to understand the company and place it in a shortlist.

Third, I would repeat the same controlled questions across engines and runs. That tells me whether the problem is stable, engine-specific, or mostly noise. With only three checks, I want each one to eliminate a different class of explanation before anyone starts proposing fixes.