Know where your brand stands in AI answers.
Compare your brand with competitors across a defined set of buyer questions. Keep the measurement scope visible and inspect the answers behind a change in visibility.
| Observation | Answers |
|---|---|
| Brand mentioned | 30 |
| Brand absent | 70 |
Missing or failed observations are separate.
- Answer-level mention rate
- 30 / 100 = 30%
- Share of tracked brand appearances
- Not calculated here
- Needs competitor counts and a stated counting rule. Multiple brands can appear in one answer.
A percentage only helps when you know what it counts.
“Share of voice” can describe different calculations. Read the definition and denominator before comparing brands, tools or reporting periods.
Answer-level mention rate
The percentage of completed answers that mention your brand. Several brands can appear in the same answer, so their mention rates do not need to add up to 100%.
Share of tracked brand appearances
Your brand’s portion of appearances across the brands being compared. This needs competitor counts and a stated counting rule. It cannot be inferred from your mention rate alone.
Build a comparison your team can explain.
A useful benchmark starts with buyer questions and ends with the evidence behind the numbers.
Choose questions that reflect buying decisions
Include customer needs, constraints and use cases: location and specialist experience for a service, or compatibility and features for a product. Keep branded questions separate from non-branded discovery.
Hold the comparison scope steady
Compare brands against the same prompts and collection conditions. Keep the engine, time window and language visible. Mark changes to the prompt portfolio, and distinguish failed observations from completed answers where the brand was absent.
Inspect the answers behind a change
Check whether the brand was recommended, compared or simply mentioned, and which sources were cited. Use the pattern to choose a research question. Repeated observations do not prove that a website change caused the movement.
Put visibility in business context
A visibility observation is not a sales result. Use it to identify questions, sources and pages worth investigating. Then examine referral visits and relevant business outcomes separately.
Frequently asked questions
What the observations mean, and where their limits are.
Is AI share of voice the same as Google ranking?
No. Google ranking concerns a result’s position for a search. AI share of voice describes a defined comparison of brand presence in an observed answer set.
Can different tools report different numbers?
Yes. Prompt selection, collection source, timing, brand matching and the calculation itself can differ. Compare the methodology before comparing the percentages.
Can I compare branded questions with non-branded discovery questions?
You can study both, but keep them separate in interpretation. A question that names your company starts with a different context from one asking for possible suppliers.
Does improved visibility prove that a website change worked?
No. It may coincide with the change, but other factors can also affect an answer. Use repeated observations and a documented baseline.
Build a benchmark your team can explain
Walk through your buyer questions and see how CiteLadder can support a more inspectable AI visibility workflow.