There is one search on Google where every competitor holds the tools to build the answer for themselves: “best SEO consultant”. Whoever ranks there is, by trade, among the best equipped people in the world to rank for it. What comes out is a one-of-a-kind SERP, a board that does not measure who is best but who is best at saying so.
We have started monitoring it continuously: who holds it, with what kind of pages, with how much earned authority, and what AI assistants answer when asked the same questions. And here is the ending in advance: when the AIs answer without searching the web, this ranking disappears. Before anything else, one necessary note. Sequel is excluded from the survey by design, because our role here is to measure, not to compete.
Best SEO consultant
How to read these numbers
Every day we record the top ten organic results on Google US across several variants of the search, always with the same parameters. What we publish is the aggregate: each site’s average position and how many variants it covers. Movements will appear week after week, and we will show them only when they exceed the natural noise of position swings. On days when Google declares a core update in progress, we flag it. And there is something anyone working with this data learns quickly: AI-generated answers change even when the same question is rephrased slightly. A single snapshot is worth little, which is why we survey every day and read trends on the series, never on a single survey.
How this world works
For those outside the industry, ranking for a search like this takes technical work every practitioner knows well, made of content built for the query and authority signals accumulated over time. It is genuine work, and those who master it prove it precisely by ranking. The question is not whether it works but what it proves, because the board rewards the ability to rank, which is one part of the craft, and says nothing about the quality of the consulting you will receive. One methodological note we will keep repeating: we are not judging individuals here, we are describing a system of incentives the industry knows well and that data finally allows us to measure.
Our surveys capture the mechanism well, with an American twist. Unlike Italy, where personal self-published rankings tend to dominate, the US top ten is held mostly by aggregators, directories, marketplaces and community threads, alongside agency pages. The pattern, though, is the same everywhere and hard to mistake for coincidence. These pages rarely rank for the value of what is written on them. What pushes them up is almost always a construction built over time: domains on which authority has been built, sites optimized for performance, and backlinks distributed too neatly to come from spontaneous citations. In the US survey the exact-match share is essentially zero, which is consistent with a SERP won by domain strength rather than by page-level link building. The aggregate numbers, refreshed with every survey, are below:
Among these numbers, the one that weighs most is the shared-sources signal: the same external sources recur across the profiles of several ranking sites at once. The more the profiles lean on the same sources, the less those links tell a story of independent authority.
What the SERP is made of
The results page is more than organic positions. Depending on the day, the variants of this search trigger Google’s AI answer, the local pack with its map, People Also Ask and paid ads. We measure them with every survey, because they tell how much attention is actually left for the ten blue links.
The heaviest block is the AI answer, which on this family of searches appears on most variants. It is field confirmation of what we wrote above, questions switch on generated answers, and it means whoever holds this ranking today also competes with the answer Google writes on its own.
What do you gain from being among the best SEO consultants in this ranking?
I can speak from direct experience here. For twelve years I worked as an independent consultant, and I entered boards like this one myself, for searches like “best web agency Rome”. We call them vanity keywords and the name is honest, but dismissing them would be superficial. Ranking first on a search your prospect can repeat in front of you is a form of social proof that always makes an impression in a negotiation, generates contacts and is a real shop window. The numbers on this page also explain why the game is so crowded, since the entry price, measured in authority to be earned, is among the lowest you can observe on a commercial query.
Then again, you also have to ask what actually gets a contract closed. A client looking for a partner on a project that matters does not stop at typing the question into Google: they ask around, cross-check references, look at results on cases similar to theirs, listen to how the industry talks about you. In my experience the share of revenue coming from vanity keywords always stayed minor, because whoever finds you in a ranking compares you with the ranking, while whoever comes through a reference has already decided half of the deal before calling.
The difference is made by what we call brand, the sum of evidence a market accumulates about you over time, the presence in the places where decisions take shape, on the web and now in AI answers, but also in professional relationships, at the events where you bring data rather than slogans, in the results clients recount to other clients. A link profile can be assembled in weeks, and the data above proves it. Becoming a brand takes years, and the traces you have seeded across the web, your real knowledge graph, are your true asset, the one that actually weighs on LLMs.
What do the AIs answer to the same question?
A Washington University study of over 55,000 searches measured that Google’s AI answers appear on roughly one search in eight, but when the search is phrased as a question the share climbs past 64%. “Who is the best SEO consultant” is exactly that kind of question, which is why we watch both arenas together, the ranking and the generated answers.
We asked the same questions to ChatGPT, Claude and Gemini, in the two conditions a user can meet them. With browsing enabled, which is how most people use these tools today, five of the ten ranking sites get mentioned, because the assistant consults the web and partly retrieves the same SERP we are observing. We then asked the same questions again from the models’ own knowledge, with no search to lean on, and the board thinned out to three, all aggregators, with none of the individual claimants mentioned. The names the assistants volunteer instead are the long-standing figures of the international SEO community, with years of content, conferences and peer recognition behind them. The mention board is below.
The distance between those two numbers is the most interesting figure on this page. As long as the assistant browses, whoever holds Google keeps a positional rent. The models’ consolidated knowledge, the one that surfaces when there is no search to lean on, rewards instead platforms and reputations built over time. These are two worlds rebalancing, and the observatory will measure in which direction.
Do AIs follow Google’s ranking?
It is the question practitioners have been asking for months, and on the US market the answer so far is that they do not, or at least not in this survey, since the sites mentioned by the AIs sit at an average position of 5.2 against 4.3 for those ignored, and the citations concentrate on aggregators regardless of where they rank. What predicts a mention here is the nature of the source rather than its position, which is a different picture from what we measure in Italy, where position and citations move together. Tracking how these two patterns evolve is one of the reasons this observatory exists.
Our own comparisons between the ranking and the generated answers show how fragile that advantage is: different engines draw from largely separate source sets, and some of the sources the AIs cite are nowhere on page one. Ranking well and being cited are two different games, and they need to be measured separately.
What the data says
This board measures presence, persistence and technique, useful information as long as it is read for what it is. If you are choosing a partner, asking Google “who is the best” hands you exactly this board, the one you have just seen. The question that actually helps is a different one, and it is specific: the best for your industry and for your problem. For the media industry, for example, we collected the criteria that matter in a guide on the best SEO consultants for publishers, covering Discover, Google News and traffic monetization.
The observatory will keep measuring, week after week: who rises, who falls, what changes when Google updates its systems, and whether the distance between the SERP and AI answers widens or closes. The data will do the rest.
Sources referenced in the text: Xu, Iqbal and Montgomery, “Measuring Google AI Overviews”, Washington University, 2026; Grossman et al., “How Generative AI Disrupts Search”, ACM SIGIR, 2026.
Frequently asked questions
