SEO metrics count positions and clicks; AI visibility metrics count mentions and citations. Rankings, impressions, and CTR still describe the classic results, but share of voice, citation frequency, and AI referral traffic describe whether engines name you in the generated answer above them.
The core difference
SEO metrics measure your position in a list of links. AI visibility metrics measure whether you are named in a generated answer. That is the whole shift in one sentence. A page can rank well and still never be mentioned by ChatGPT, and a page can be cited constantly while ranking modestly.
So the old numbers and the new numbers are not rivals; they measure different layers of the same search. You need both, but you have to know which is telling you what.
Which SEO metrics still matter
Plenty of classic SEO metrics keep their value, because being indexed and readable is still the entry ticket to AI answers.
- Indexation: if you are not in the index, you cannot be cited
- Crawlability and page health: engines have to read you
- Backlinks and authority: still a strong trust signal
- Rankings and impressions: useful context, no longer the headline
The connection between the two worlds is covered in GEO vs SEO; the short version is that good SEO is the foundation AI visibility builds on.
Why rankings stopped being the headline
Rankings once told you almost everything. Now they tell you less. Google AI Overviews appear on a large and growing share of searches, and Ahrefs found the top organic result loses roughly a third of its clicks when an AI Overview shows. A number one position under a generated answer captures far fewer people than it used to.
That is why leading with rank is misleading today. The click you used to win is increasingly absorbed by the answer above your listing.
The new metrics that replace them
Three new numbers carry the load. Share of voice is how often you are named in category answers. Citation frequency is how often your own pages are the source an engine quotes. AI referral traffic is the visitors those answers send you.
These sit at the center of the GEO KPIs that actually matter. They measure presence in the answer, authority as a source, and the traffic that results, which is exactly what rankings and CTR used to approximate.
How you measure each
The measurement methods differ, and that trips people up. SEO metrics come from tools wired into the index: Search Console, analytics, backlink checkers. AI visibility metrics mostly come from running a fixed prompt set by hand across the engines and recording what you see.
Referral traffic is the exception, since it lands in your analytics. Capturing it cleanly is its own small discipline, covered in how to track AI referral traffic, because AI referrers do not always announce themselves clearly.
Reading them together
Use the two sets side by side. Strong SEO metrics with weak AI visibility usually means your content ranks but is not structured or corroborated enough to be quoted. Strong AI visibility with thin traffic means you are named but the answers are not sending clicks yet, which is normal early on.
Neither set alone tells the full story. Read together, they show whether you are indexed, whether you are cited, and whether any of it is bringing people to you. That is the complete measurement picture for search today.
Building one blended scorecard
The practical move is to stop running two separate reports and build one scorecard that blends the layers. Buyers experience search as a single thing now, part links and part generated answers, so your measurement should reflect that unified reality.
Group the numbers by what they tell you. Foundation metrics confirm you can be found at all: indexation, crawlability, and authority. Presence metrics show whether you appear in the answer: share of voice and mention rate against competitors. Source metrics show whether you are the one being quoted: citation frequency of your own pages. Outcome metrics show whether any of it brings people to you: organic clicks plus AI referral traffic in your analytics.
Read top to bottom and the diagnosis writes itself. Strong foundation but weak presence means your content is found but not quote-ready. Strong presence but weak outcomes means you are named but the answers satisfy users without a click, which is common and often fine. One blended scorecard, reviewed on a cadence, keeps the whole search picture in view instead of forcing you to reconcile two disconnected stories.
Questions people ask
No. Indexation, crawlability, page health, and backlinks all still matter, because being readable and trusted is the entry ticket to AI answers. What changed is that rankings and click-through rate are no longer the headline, since generated answers absorb clicks above the traditional results. Keep the SEO metrics as foundation and context, and lead your reporting with the newer visibility numbers.
Three carry most of the weight: share of voice, meaning how often you are named in category answers; citation frequency, meaning how often your own pages are the quoted source; and AI referral traffic, meaning visitors arriving from engines like ChatGPT and Perplexity. The first two come from running a fixed prompt set by hand; the third comes from your analytics.
Yes, and it is common. Ranking measures position in a list of links; citation measures whether an engine quotes you in its generated answer. A page can rank strongly yet be too vague, unstructured, or poorly corroborated to be quoted. That gap is exactly why you need AI visibility metrics alongside your existing SEO metrics rather than instead of them.
Want this done for you?
Want to know if AI is recommending you or your competitor? Get an AI visibility audit from Acromatico.
Get a free AI Visibility Audit →