Real measurement notes with fixed questions, provider provenance, failed-run handling and explicit limits—not anonymous uplift claims.
HowAICite sells software in the category and produced these measurement notes. Customer relationships are never implied. A missing or failed observation is disclosed and excluded—not silently converted into a negative result.
Ten US-English buyer questions observed twice through Perplexity Sonar API. No own-brand mention or own-domain citation was observed.
Ten non-branded AI visibility questions across four provider APIs. HowAICite appeared in none; Profound appeared in 16 and Peec AI in 14 valid answers.
Six frozen buyer questions were attempted across four provider APIs. The note separates completion, brand mention and own-domain citation denominators.
The exact buyer questions and whether they were selected, frozen or weighted.
Provider, model when available, requested market, language and collection time.
Attempted, completed, failed, missing and excluded observations remain distinct.
Mention, recommendation and citation are separate; none automatically proves revenue or causality.
The October competitor benchmark is the current before-state. After indexing and independent distribution, HowAICite will repeat the same question, provider, market and language combinations. Any change will be published beside—not in place of—the baseline.
Read the measurement method →Run a free battle scan on your domain. It asks the AIs real buyer questions and shows which rivals they name. No signup.