FAQ & troubleshooting

My dashboard is empty

In order of likelihood:

1. No prompts yet — add them in the Prompts tab.
2. Prompts added, first cycle not run — collection is scheduled, not instant. See Data freshness.
3. The period predates your data — widen it to 30 days.
4. A filter is excluding everything — check the brand, group and assistant filters.

The site audits under the Report tab work immediately and need none of the above.

Visibility dropped and nothing changed on our side

• Check the period. A 1-day view is a sample. Compare 7- or 30-day windows before concluding anything.
• Check whether prompts changed. Adding prompts you do not yet appear in lowers visibility without anything getting worse.
• Check the assistant filter. Enabling an assistant where you are weaker changes the pooled number.
• Accept some variance. The same prompt can produce different brands on consecutive days — see Data freshness.

We are mentioned in the answers but visibility says otherwise

Almost always name matching. A mention counts only if the form in the answer matches a configured name or variation.

Open the answer, read exactly how your brand is written, and add that form as a variation in Brands. Common culprits: abbreviations, legal suffixes, transliterations and misspellings.

Our visibility is near zero everywhere

Work down the supply chain:

1. Robots.txt audit — are AI crawlers blocked?
2. Bots tab — do they actually visit?
3. Technical checks — is the page blank without JavaScript?
4. Sources tab — which domains win these answers instead?

A blocked crawler or a client-rendered page explains more zero-visibility cases than content quality does.

Share of Voice fell but visibility rose

Expected, and not a contradiction. Visibility counts answers; share of voice counts mentions and is measured against every tracked brand. Appearing in more answers while being named more briefly — or a competitor being named far more — produces exactly this. See Share of Voice.

Also check whether a competitor was added: a new brand in the set lowers everyone's share without anything changing in the answers.

Position says 1.0 but it is marked "not reliable"

The average rests on too few appearances to be meaningful. Read Visibility instead until the brand appears more often — see Position.

Sentiment shows nothing for our brand

There were no mentions to score in the period. A score is shown only when at least one answer mentioning the brand was scored — showing 50 with no data would read as "neutral coverage", which would be false.

Two assistants disagree completely

Normal. Assistants rely on different sources and phrase recommendations differently. Filter Sources to the weaker assistant to see which domains it uses — the gap is usually a source gap.

No crawler data appears

1. Confirm the integration is deployed on the hostname you are measuring.
2. Confirm the ingest key is exactly the one that was generated.
3. Check your platform's logs for errors from the forwarding code.
4. Remember your own visits do not appear — only identified crawlers are counted.

See Bot analytics.

An audit finding does not match our page

Findings are written to describe the page in front of them, so a mismatch is worth reporting rather than working around. Re-run the audit first: results persist, and you may be reading findings from before your last deployment.

Our trial ended — is our data gone?

No. Collection stops; prompts, competitors, metrics and audit results remain. See Trial and paid plans.

How long until changes we make show up?

Publishing a change is the start of a chain: crawlers have to fetch it, assistants have to incorporate it, and answers have to be collected. Expect weeks rather than days, and watch Bots and Sources for the earlier signals that something moved.