On this page
Key takeaways
- There are two numbers. The AEO score (0-100) shows how ready a page is for AI assistants. AI visibility shows how often ChatGPT, Gemini and Perplexity actually cite you.
- A "good" AI visibility score is relative: it’s good if you’re cited more often than the competitors buyers compare you with.
- A good AEO score is one with no big gaps in the heaviest areas: content answerability (35%) and structured data (25%).
- A high AEO score with low AI visibility usually means competitors answer the same questions better.
- After fixes, AI citations usually take about 6 to 8 weeks to change, so measure again then.
What is a good AI visibility score? There isn’t one number that counts as good for every business. AI visibility is a share of answers: how often ChatGPT, Gemini and Perplexity cite your website when buyers ask the questions you want to win. A good score is one that puts you ahead of the competitors buyers compare you with, and shows improvement over time. There’s no universal target, because AI visibility is relative to your market and the questions being asked. A good AEO score, the readiness number from 0 to 100, is one with no major gaps in the areas that carry the most weight.
Below: why it matters, what each number measures, how to tell whether yours is good, and what to fix first.
Why AI visibility matters
When buyers use ChatGPT, Gemini or Perplexity to research vendors, the answer usually names only a handful of companies and cites only a few websites. If your business isn’t one of them, you may never make the buyer’s shortlist, even if you rank well in Google.
So the score itself isn’t the goal. A good AI visibility score means your company is included in the answers buyers actually read when they compare products, services and vendors: more chances to be recommended, fewer recommendations going to competitors, and more of the enquiries that follow.
This matters most for businesses that are chosen after research, such as software development firms, IT service providers, consultants, managed service providers and other B2B companies. Their websites often rank reasonably well in search, but that doesn’t mean AI assistants cite them.
Two numbers, two different questions
Most AI visibility reports, including ours, give you two numbers. They answer different questions, so read them separately.
| AEO score | AI visibility | |
|---|---|---|
| The question it answers | Is the page ready for AI assistants to find, read, understand and quote? | Do AI assistants actually cite it? |
| Scale | 0 to 100 | A percentage of answers, with the count behind it (for example "15%, 20 of 135") |
| What moves it | Changes to your own page | Your page, and what competitors publish |
| What it’s for | Knowing what to fix | Knowing whether the fixes are working |
Think of a shop. The AEO score measures whether customers can find the shop and understand what it sells. AI visibility measures how often customers actually choose to walk in.
What is a good AEO score?
The AEO score combines five areas. Each area is a set of checks, and each check is worth a share of that area’s points.
| Area | Weight | What it checks |
|---|---|---|
| Content answerability | 35% | Question-style headings, a direct answer at the start of each section, lists and tables for steps and comparisons, specific facts |
| Structured data | 25% | JSON-LD schema (Organization, FAQ, Article, Product, Breadcrumb) that’s present, valid and matches what the page shows |
| Crawl access | 20% | AI crawlers allowed in robots.txt, main content in the HTML without JavaScript, canonical link, sitemap, llms.txt |
| Technical health | 10% | Google PageSpeed Insights scores |
| Trust signals | 10% | Who is behind the page, a named author, sources for claims, a visible "last updated" date |
The full method, including how PageSpeed results are turned into points, is on our methodology page.
How to read your score. As a rough guide:
- 80 to 100: the page is in good shape. Remaining fixes are usually small, and your AI visibility depends more on how well you answer specific questions than on technical problems.
- 60 to 79: solid, with clear gaps. Usually missing schema, generic headings, or no direct answers near the top of sections.
- Below 60: something basic is likely in the way, such as AI crawlers blocked in robots.txt, content that only appears after JavaScript runs, or no structured data at all. Fix these first: they stop AI assistants from reading the page, however good the content is.
Two more ways to judge it:
- Look at the area scores, not just the total. A 75 with content answerability at half marks is a bigger problem than a 75 that lost points on page speed, because content carries 35% of the score and speed 10%.
- Compare with your competitors’ pages. A 70 is a good score if the pages AI assistants cite instead of yours score 55. It’s a weak one if they score 90.
What is a good AI visibility score?
AI visibility is measured by asking AI assistants real buyer questions and recording which websites each answer cites. Because answers change every time, it’s reported as a rate over many answers. Our checker asks 15 questions, 3 times each, on ChatGPT, Gemini and Perplexity: 135 answers for one page.
You’ll usually see three figures:
- Site citation rate: the share of answers that cite any page of your site.
- Page citation rate: the share that cite the page you checked.
- Share of voice: your citations as a share of all citations in all answers. Each answer cites several sites, so this is usually much lower than the citation rate.
So what’s good?
- Being ahead of your competitors. This is the number that matters. Buyers see a short list of names in each AI answer; if a competitor is cited in 30% of answers and you’re cited in 10%, they’re on that list three times as often.
- Winning the questions closest to a sale. Being cited for "Who’s the best [service] in [city]?" matters more than being cited for a general definition. Read the results per question, not just the total.
- Trending up between scans. Rates move from scan to scan because AI answers vary. Look for a change that holds across several questions, not one jump.
- Any citations at all, if you’re starting from zero. Many sites aren’t cited in any answer for generic buyer questions. Getting into some answers is the first milestone.
There’s no universal "good" percentage, because markets differ so much. In a crowded market with many strong competitors, a fairly low rate can still put you ahead of most of them. In a niche market, the leading company may be cited in a much larger share of answers.
An example. Suppose you run a precision machine shop in Michigan. You check 15 questions a buyer might ask, such as:
- "Best CNC machining company in Michigan"
- "Precision machine shop near Detroit for prototype parts"
- "Contract manufacturer for low-volume automotive parts"
Asked 3 times each on ChatGPT, Gemini and Perplexity, those questions produce 135 answers. If your site is cited in 27 of them, your citation rate is 20%. If your closest competitor is cited in 45, theirs is 33%: they have more AI visibility, even if your AEO score is higher. That’s the gap to close, and the per-question results show where.
(This example is made up to show the calculation. It isn’t a benchmark.)
Only generic questions count, the ones a buyer asks before they know your name. A question like "What does [your company] offer?" almost always cites your own site, so it’s reported separately and left out of the score.
Why a high AEO score can come with low AI visibility
This is common, and it’s the reason both numbers are worth having. Readiness removes the reasons AI assistants skip a page, but they still choose between pages. If a competitor’s page answers the same question more directly, or is better known for the topic, it’s cited instead.
When that happens, look at the questions you lose. The competitor page that wins each one shows what AI assistants preferred: often a more direct answer, a comparison table, or specific numbers your page doesn’t give.
The opposite also happens: a page with a weak AEO score that’s still cited, usually because the site is well known for the topic. Fixing its readiness gaps protects those citations.
What to fix first
- Anything that blocks reading: AI crawlers blocked in robots.txt, content that needs JavaScript, a page that redirects or isn’t in the sitemap. Usually quick fixes for a developer.
- Content answerability: headings that ask the questions buyers ask, a one- or two-sentence answer straight under each heading, and specific facts instead of general claims.
- Structured data: Organization, FAQ and Article schema that matches what the page shows.
- Trust signals: a named author, a visible "last updated" date, and sources for claims.
- Page speed and accessibility last, unless they’re badly off.
Then wait. After fixes, AI citations usually take about 6 to 8 weeks to change, because AI assistants have to find and re-read the page. Check the same page again then and compare.
How to check your scores
You can check AI visibility by hand: ask ChatGPT, Gemini and Perplexity your buyers’ questions and write down who’s cited. It works for a handful of questions, but it gets slow quickly, and it doesn’t tell you why a page isn’t cited.
Our free AI visibility checker does both: it gives you an AEO score and tests 3 questions on screen with no signup, and the full report (15 questions, 3 competitors and a ranked list of fixes) arrives by email in a few minutes.
Frequently asked questions
Is there a standard AEO score everyone uses?
No. Each tool calculates its own score with its own checks and weights, so a 70 in one tool isn’t the same as a 70 in another. Compare scores from the same tool, and look at what the tool checks. Ours publishes its methodology.
Can a high AEO score guarantee AI citations?
No. No AI company publishes how it chooses sources, so no tool can guarantee citations. A high score means the page removes the common reasons AI assistants skip it. AI visibility shows whether that’s turning into citations.
Can I have strong SEO but poor AI visibility?
Yes. SEO measures how well a page performs in search engines. AI visibility measures how often AI assistants cite it when answering questions. A page can rank well in Google and still be cited rarely, if competitors give more direct answers or stronger supporting content.
Why does my AI visibility change between scans?
AI assistants write each answer fresh, and the pages they draw on change too. Asking each question several times and reporting a rate reduces the noise, but some movement is normal. Look for changes that hold across several questions.
How long does it take to improve AI visibility?
After you make fixes, AI citations usually take about 6 to 8 weeks to change. Check the same page again then, and compare it with your earlier results.
Does improving my AEO score help SEO too?
Usually, yes. Most fixes, such as clear headings, direct answers, valid schema and crawlable content, are also good SEO practice. They can help with Google’s featured snippets and AI Overviews as well.
Do the results match what I see in the ChatGPT app?
Not exactly. Our checker uses the engines’ official APIs with web search turned on. Answers in the apps can differ with a person’s location, chat history or settings. That’s why the results are rates over many answers, not a single check.
Not sure where your pages stand? Check your AI visibility for free, or talk to us if you’d like help with the fixes.
Join the discussion
No comments yet.
Leave a Reply