What is an AI visibility scorecard?
An AI visibility scorecard is a structured baseline of your brand’s presence across AI-generated answers, scored against named competitors for the real questions your buyers ask.
It is not a vanity score. Each figure in it maps to something you can act on. NeuralGen’s version is free, delivered within 48 hours, and covers 20 real buying prompts run across five AI engines. For the metric behind the report, see our guide to AI visibility, and for the offer itself, visit the AI Visibility Scorecard page.
The name is deliberately borrowed from a familiar format. A credit scorecard turns a fuzzy question — “am I a good credit risk?” — into a small set of concrete numbers you can act on. An AI visibility scorecard does the same for a different fuzzy question: “would an AI assistant recommend my business?” Both replace a guess with a documented baseline.
Most businesses requesting one have already had an unsettling experience: they asked ChatGPT or Perplexity a question about their own sector, expecting to see their name, and either didn’t appear or saw a competitor named instead. A scorecard turns that one anecdotal moment into a structured, repeatable check across the questions that actually matter to their buyers, not just the one prompt they happened to try.
That single-prompt experience is a reasonable starting alarm, but it is a poor basis for decisions on its own — the exact wording of a question can change the answer, so one good or bad result does not tell you much about your overall position. A scorecard exists precisely to replace that anecdote with a number built from enough prompts and enough engines to be a fair read.
What does an AI visibility scorecard measure?
A scorecard measures five things for each buying prompt: presence, prominence, sentiment, citations and the roll-up share-of-voice figure.
- Presence: whether your brand appears in each answer, for each of the 20 prompts tested.
- Prominence: how featured you are when you do appear — named first, mentioned in passing, or listed alongside several competitors.
- Sentiment: how the AI describes you, including whether the description is accurate and current.
- Citations: which sources the AI credits for its answer, so you know exactly where to focus corroboration efforts.
- Share of voice: the roll-up figure versus competitors, expressed as a single trackable percentage.
Each metric on its own tells only part of the story. A brand can be present in every answer yet always mentioned last, which a presence figure alone would not reveal — that’s what prominence is for. A brand can be present and prominent yet described with an outdated price or a service it no longer offers, which is what the sentiment and accuracy check is for. Reading all five together, rather than fixating on one, is what makes the report useful rather than reassuring-but-hollow. Our generative engine optimisation guide explains how each of these components maps to the fixes that move it.
How much does an AI visibility scorecard cost?
NeuralGen’s AI Visibility Scorecard is free, delivered within 48 hours, covering 20 real buying prompts across five AI engines against named competitors.
For those who want the fuller diagnosis, the paid AI Visibility Audit is £1,800, fixed, delivered in five working days — it adds entity analysis, a citation-source map, llms.txt and schema fixes, and a 90-day roadmap. The scorecard is the free starting point; the audit is the paid deep dive. Full detail is on our AI SEO agency page.
There is no hidden upsell built into the scorecard itself — you get the full 20-prompt, five-engine result regardless of whether you go further. The audit exists for businesses that already know they have a gap and want the root-cause analysis and a fixed sequence of work, rather than working through the pillar guides themselves.
How do you read your AI visibility scorecard?
Start with your headline share-of-voice figure, then work through the gaps it exposes, in this order:
- Start with share of voice — your headline position versus competitors, and the number you will track from month to month.
- Check presence gaps — the prompts where you are absent are your biggest opportunities, because absence is the easiest problem to make progress on quickly.
- Read sentiment — being mentioned negatively or inaccurately is its own problem to fix, separate from simple absence, and sometimes more urgent.
- Review citations — the sources AI trusts show you exactly where you need to appear next; if a competitor’s directory listing keeps getting cited and yours doesn’t exist, that’s a concrete, fixable gap.
- Note prominence — being buried in an answer is weaker than being named first, even if you technically “appear”, so don’t treat every mention as equally good news.
- Set a re-measure date — one month is a sensible cadence to see whether fixes are working, since AI engines refresh their sources on their own schedule rather than instantly.
See our AI visibility guide for how each metric behaves over time, and our llms.txt guide for one of the concrete fixes a scorecard often surfaces.
Resist the urge to fix everything the first week. A scorecard usually surfaces more issues than you can address at once, and trying to fix them all simultaneously makes it hard to tell which change actually moved the number next time you measure. Pick the two or three biggest presence gaps, fix those properly, and let the next measurement tell you whether it worked before moving on.
What should you do after reading your scorecard?
Prioritise the biggest presence gaps and the most-cited sources you are missing, fix your entity and content structure, then re-measure.
You have two paths from here: do it yourself using our pillar guides, or engage NeuralGen to run the fixes and the measurement for you. Either way, the honest line holds: this is measurement and method, not a guaranteed citation. If you’d like it done for you, our AI SEO agency service is the next step.
A sensible first pass looks like this: fix your entity details everywhere they appear, rewrite the one or two pages most relevant to your weakest prompts so the key facts sit in the first sentence, and chase corroboration — reviews, directory listings, press — for the sources your scorecard shows you are missing from. That is usually two to three weeks of focused work rather than a major project, and it is the same sequence whether you do it yourself or ask us to run it.