Reading Your Results

Each Brand Alignment scan shows how accurately AI engines represent your brand, fact by fact, and how they position you. Every result leads back to the answers behind it.

Filters

The filters at the top apply to the whole page:

• Scan: the scan to show. The latest finished scan is selected by default. Compare shows the change from the previous scan.
• Models and Markets: the engines and countries to include.

Results combine the answers to the scan’s own questions with the sampled answers from your tracked prompts that mention your brand.

How Accurately Is Your Brand Represented?

The cards under this heading on Overview count the answers that address your facts:

Each assessed answer counts once. When an answer both supports and contradicts your facts, it counts as Contradicted. Unresolved identity is measured on the scan questions only. Answers that take no clear position on a fact are marked Inconclusive: they stay available to inspect but are left out of these rates.

Select a card to see the answers it counts.

Your Brand Facts

In Setup › Facts, each fact is listed by category with the number of answers in each outcome for the selected scan. A fact added since that scan shows Pending next scan. Open a fact to see:

• Alignment over time: how its outcome moved from scan to scan
• Results by model: the outcome on each engine
• Answers assessed against this fact: each answer, with the statements that conflict with the fact

Statements and Sources

Expand an answer to see the statements that conflict with your facts. Each statement is labeled:

• Contradicts: the statement is incompatible with your fact
• Partly contradicts: the statement supports part of your fact and conflicts with another part

Each label comes with the exact quote from the answer, the fact it was checked against and, under Cited for this statement, the source the engine attached to that statement, where it attached one. Read full answer shows the whole answer and all of its cited sources.

A cited source shows where the engine says its information came from. It is the best place to start a fix, but it is not proof of where the engine learned the statement. Fixing What AI Gets Wrong explains how Asky handles both cases.

How AI Positions You

This section summarizes how engines describe your brand and compares it with the positioning statement in your Brand Profile. It combines the scan’s brand-description answers with tracked answers that mention your brand. Select an engine to see its own view.

At a glance lists each part of your positioning with what the engines did with it:

Each scan keeps the positioning statement it compared against, so editing your Brand Profile does not change past results.

Attributes and Sentiment

Attributes are the traits engines associate with your brand, such as Easy to use or Expensive. For each attribute you see how many answers associate it with your brand, on which engines, and with what sentiment: Positive, Mixed, Negative or Neutral. Open an attribute to read the quotes behind it. Manage the list in Setup › Attributes.

Trends

The history card shows Alignment, Sentiment and Attributes over time, for the filters you chose. Lines combine all selected engines by default; switch on Compare models to see each engine on its own.

Cited Sources

The Sources linked to contradictions card on Overview shows the five pages most often cited for a statement that conflicts with your facts. The Cited sources tab lists every page engines cite in these answers, sorted by how many contradictions each is linked to.

Each source is marked Internal when it is on one of your domains and External otherwise. If Asky gets that wrong, for example for a help center on another domain, change it on the source’s row. The setting covers the whole host and only affects Brand Alignment.

Next Steps

• Fix what engines get wrong with Fixing What AI Gets Wrong
• Tighten the facts you check in Facts and Markets