AEO Competitive Visibility Scorecard
Judge AI visibility through share of voice, sentiment, and prompt-level context
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 6
- Confidence
- 98%
The AEO Competitive Visibility Scorecard evaluates a brand across three connected levels. First, prompt-level analysis records what each answer engine says in response to the tracked customer questions. Second, aggregate visibility shows how often the brand appears across that prompt set and how the result changes over time. Third, share of voice compares those mentions with named competitors, while sentiment distinguishes helpful recommendations from negative discussion. This prevents a marketer from celebrating a visibility percentage in isolation: the same percentage may be strong or weak depending on competitors, and frequent mentions may still damage the brand when their tone is unfavorable. Running the scorecard repeatedly also exposes volatility and provides a baseline against which later optimization work can be judged.
Origin
Extracted from Marketing Against The Grain during a HubSpot AEO walkthrough using Dell, Lenovo, HP, and IBM as the competitive set.
Core principles
- 01Visibility has meaning only when compared with relevant competitors.
- 02Mention volume and mention quality are separate dimensions.
- 03Prompt-level results explain aggregate visibility scores.
- 04Recurring measurement is necessary because answer-engine results fluctuate.
How to run it
- 1
Normalize brand identities
Add the principal brand name and meaningful variations so mentions can be recognized consistently across answer-engine responses.
Pro tip Include common abbreviations and legitimate naming variants.
Watch out Poor identity coverage causes genuine mentions to be missed.
- 2
Define the competitor set
List the alternatives most likely to be recommended alongside or instead of the brand. Choose competitors from the customer's perspective, not merely an internal corporate list.
Pro tip Include emerging alternatives when answer engines recommend them frequently.
Watch out Looking only at the focal brand cannot reveal lost recommendation share.
- 3
Collect recurring responses
Send the tracked prompt set to the selected answer engines on a regular cadence and retain the responses. Recurring collection captures normal fluctuations and material changes.
Pro tip Use a consistent prompt set for trend comparisons.
Watch out A single snapshot may reflect answer-engine volatility rather than durable performance.
- 4
Measure visibility
Calculate how often the brand appears across all prompts, then inspect the individual prompts responsible for the aggregate result.
Pro tip Separate prompts with brand mentions from prompts where no brand is recommended.
Watch out An aggregate score can conceal strategically important prompt-level gaps.
- 5
Calculate share of voice
Compare the focal brand's mentions with competitor mentions across prompts that mention brands. Use the relative shares to interpret whether the visibility level is competitively strong.
Pro tip Watch the direction of competitor movement, not only the current rank.
Watch out A brand can remain first while a competitor rapidly closes the gap.
- 6
Assess sentiment
Classify whether mentions are positive, negative, or neutral. Combine this result with mention frequency before judging performance.
Pro tip Review the underlying wording for high-value or unusually negative prompts.
Watch out High visibility built from unfavorable commentary is not a marketing win.
In the wild
The episode's demonstration shows Dell with 52.8% of brand mentions and Lenovo with 42%. Dell remains first, but Lenovo is close enough to warrant attention. Sentiment analysis is then added to determine whether Dell's numerous mentions are actually favorable rather than assuming all visibility has equal value.
→ The marketer interprets visibility through competitive position and mention quality instead of relying on a standalone percentage.
Common mistakes
Reading visibility in a vacuum
A brand's mention rate says little until it is compared with the competitors receiving the remaining recommendations.
Counting all mentions as wins
Negative mentions can increase visibility while weakening customer preference, so sentiment must be reviewed separately.
Overreacting to one day's result
Answer-engine outputs fluctuate, making recurring observations more reliable than isolated snapshots.
Is it for you?
Best for
Brands that need a repeatable scorecard for comparing their answer-engine presence with the alternatives recommended to customers.
Not ideal for
Brands with no defined competitor set or too little prompt coverage to make relative visibility metrics meaningful.
From the transcript
“Well, in a vacuum, that doesn't necessarily say anything. But I think the big question is, how are we doing against our competitors?”
“And so one of the key measurements to look at that is share a voice. How are we actually doing against our competitors?”
“The other dimension you want to look at is actually sentiment analysis, which will actually tell us, okay, I'm being mentioned a lot, but are…”
From the episode
How Brands Win in ChatGPT Search