How scoring works
A report asks every prompt of each provider you picked, taking as many samples as you configured. Each answer is then judged: was the brand mentioned, actually recommended, the answer's top pick, cited via its own website, or framed negatively — and which competitors showed up.
Most rates below are shares of the report's scored answers; the two share metrics instead divide by every appearance or recommendation across your brand and its competitors. Either way, more scored answers — more prompts, providers, or samples — means steadier rates.
The rates
- Mention rate
- The share of scored answers that mention your brand at all — the baseline visibility question: do you show up?
- Recommendation rate
- The share of scored answers that actually recommend your brand. Being named in passing doesn't count.
- Primary pick rate
- The share of scored answers where your brand is the answer's single top pick.
- Citation rate
- The share of scored answers that cite your own website as a source, matched against the project's domains as they were when the report was created — later edits don't rewrite old reports.
- Negative sentiment
- The share of scored answers that frame your brand negatively. The one rate where lower is better.
- Share of voice
- Of every brand appearance in the answers — yours plus your tracked competitors' — the share that is yours. Blank when no one appears at all.
- Share of recommendations
- Of every recommendation the answers handed out, the share that went to your brand. Being criticized by name lifts share of voice, but not this.
Why numbers move between reports
AI assistants don't answer the same question the same way twice, so some movement between runs is ordinary noise — especially on small runs with few scored answers.
Repeat sampling is the countermeasure: more samples per prompt mean each rate rests on more answers, so real shifts stand out from randomness. Adding or removing prompts also moves the aggregate rates on its own; the report calls that out when it happens.
When a change is significant
Each report is compared against the project's previous scored report. A shift is only flagged — on the report page and in alert emails — when it moves at least 15 percentage points and is large enough, given how many answers both runs scored, that it's unlikely to be run-to-run randomness.
Smaller moves still show as the "vs previous" deltas beside each metric; they're just not treated as news.