SEO ranks pages; AI visibility ranks mentions
Search engine optimisation is built around a simple unit: the page. A search engine reads the open web, scores every page it can find for a given query, and returns a ranked list. The page that wins is the one the engine judges most relevant for that query. SEO work — title tags, internal links, content depth, backlinks, page speed — is all in service of moving one specific URL up that ranked list.
AI visibility is built around a different unit: the mention. When a buyer asks ChatGPT or Claude "what's the best tool for X?", the AI doesn't return a list of pages. It returns a synthesised paragraph that names two or three brands. Your goal isn't to outrank a page; it's to be named in that paragraph. The measurement is whether your brand surfaces at all, and how often, across the kinds of questions your buyers actually ask.
Both disciplines share the strategic goal — be where buyers look. The mechanics are different enough that an SEO playbook doesn't translate directly. Beating the page above you in Google doesn't change whether ChatGPT mentions you.
The data sources are different
Search engines crawl the open web on a roughly continuous cadence. A page you ship today can be indexed and ranking within days, sometimes hours. The data behind a search result is fresh, and SEO tools can read the same index — keyword rankings, backlinks, organic traffic — from the same crawler-driven source the engine itself uses.
AI assistants work from a mix of training data and retrieval. Training data is a snapshot from some point in the past; the cadence is months, not hours. Retrieval — the ability to fetch fresh information at query time — varies by provider, by query, and even by user. A new page on your site might be visible to an AI assistant tomorrow, or it might not influence its answers for many months.
This has practical consequences. The signals that move SEO results (a new backlink from a high-authority site, a refreshed page with better keyword targeting) don't immediately show up in AI answers. The signals that do shift AI answers — being mentioned across many third-party sources, having a presence in directories and comparison content that crawlers feed back into training pipelines — operate on a different timescale and look more like reputation work than page work.
The measurement is different
An SEO score is a position in a ranked list. You can run the same query, get the same set of ten links back in the same order, and watch a single number move. Two queries from two people on two devices in the same week produce comparable rankings.
An AI visibility score has to deal with a property search results don't have: the same question can produce different answers from the same model run twice. The model can differ between runs even when the question and the moment are identical. So a defensible reading can't be built on a single question asked once.
Lumialo handles this by asking each question multiple times — three runs per provider on the paid plan — and treating a brand as mentioned for that question only when it surfaces in at least two of the three runs. That stricter-than-one rule is the difference between a mention you can repeat tomorrow and a one-off that wouldn't appear if you'd asked an hour later. The score itself is straightforward: the share of analysed questions where your brand was named, expressed as a percentage. No position weighting, no sentiment weighting in the headline number — those are recorded alongside but kept out of the score so it stays easy to explain.
What stays the same: content quality matters
It would be a mistake to read the differences above as "SEO doesn't matter for AI". Content quality, brand mentions, and a clear web presence still matter — they're just a smaller share of the picture than they are in classic search.
Both disciplines reward the same brand behaviour: being recognisably the same company, talked about by the same kinds of sources, with a consistent description of what you do. A brand that's well-positioned in search tends to also have most of the inputs an AI model needs to describe it accurately. The reverse isn't always true — a brand can rank well for a few high-traffic queries and still be invisible to ChatGPT — but the overlap is real.
The honest framing is that AI visibility and SEO are two separate readings of the same brand. You can be strong on one and weak on the other. You can be strong on both. The first step is to measure both and stop assuming a search ranking is also an AI ranking.
Where Lumialo fits
Lumialo measures the second reading. It runs the buyer questions you care about against ChatGPT and Claude, repeatedly, and gives you one defensible percentage — your AI Visibility Score — plus the list of questions where you were absent and the competitors who surfaced instead.
The methodology page walks the score formula, the providers, the runs per question, and the evidence rules for recommendations end-to-end, with no marketing wrapper.
If you already invest in SEO, AI visibility is the missing reading on the other half of the buyer journey. If you don't yet have an SEO programme, AI visibility is one of the cheapest ways to find out where you stand on a discovery channel that's growing fast and that most teams aren't measuring at all.