Brand perception

Head to Head

Compare vs. competitor

A controlled comparison restricted to the answers naming both your brand and one competitor you choose: coverage, who is named first, sentiment and role mix, and, on Starter, Growth and Pro, the same reason and category rates as Brand reasons, all computed on that shared population. The competitor side needs a tracked competitor, included on Starter, Growth and Pro and during the product trial.

• Only you, only them, both, neither: a coverage bar splits every analysed answer four ways before the comparison narrows to the answers naming both.
• Ties counted as ties: when both brands are first named at the same point in an answer, it is a tie, credited to neither side.
• Paired evidence: open any row of the per-prompt table to read both brands' saved quotes, sentiment and role from the same answer.
• No winner score: the two sides are set next to each other from saved evidence, with no composite score and no claim about why they differ.

Weekly Study

Objections

On Starter, Growth and Pro, a study asks each chat engine on your plan, and ChatGPT (app) (not Google AI Overviews, AI Mode or Gemini (app)), why a buyer might not choose your brand and each active tracked competitor. It runs every week, and owners and editors can also run it on demand, at most once a day. Each objection is ranked by where it appears in the answer, grouped with the same objection from other engines and brands, and shown with the quote from the answer and, where the engine records where it placed its citations, the sources it cited there. See which objections engines list first, whether competitors face the same ones, and what changed since the last comparable study. The question asks for downsides, so most brands get some: compare them rather than reading them as a verdict. Studies are separate from your tracked prompts and change no visibility metric.

• Prominence across engines: each engine scores an objection by how early it listed it, and an engine that answered without raising it scores 0, so prominence is the average over every engine that answered.
• Category or just you: on your own brand, see how many competitors had the same objection raised about them.
• Sources linked by our code: a source is tied to an objection only where the engine placed a citation on the quote or right after it, never by asking a model to match them.

Accuracy

Fact check

On Starter, Growth and Pro, keep a list of approved facts about your brand, such as pricing, founding, availability and policies, typed by your team or drafted from your own site and approved before use. A weekly study asks each chat engine on your plan, and ChatGPT (app) (not Google AI Overviews, AI Mode or Gemini (app)), about each category you have facts in, except Other, and the answers to the tracked prompts you choose are checked as they arrive. An AI model judges whether each answer supports or contradicts each fact; our code keeps a finding only when its quote is found in the answer and its fact is one of yours, and links the sources the engine cited there. Contradictions wait in a review queue, where your team confirms or overturns them or marks your own fact as out of date, and raise an alert. See accuracy per fact and per engine, and what changed since the last study.

• Questions that do not lead: the study asks about your brand by its name and domain, never with the text of a fact, so the engine is not steered toward the answer you want.
• Checked prompts: choose up to 10 tracked prompts, and their new answers are checked against your facts as they arrive, up to 200 checks a day, with any answer skipped by that limit counted on screen.
• A verdict your team owns: mark a contradiction confirmed wrong, mark that the answer agrees with the fact, mark your own fact out of date, or mark it not relevant, and add a note.