European AEO/GEO Services Market: Public Agency Offers Across 30 Countries

The criteria below were fixed before collection began. They divide the 1,051 records into three classes.

Within the active group, the offer was confirmed in 625 records. In seven it remained unknown: the site was reachable, but the offer could not be confirmed to the evidence standard.

Throughout the study, a value that could not be established is recorded as unknown — never as a negative finding. This distinction carries most of the weight in the pricing and service-stage results below.

Discovery ran in two channels, country by country, in three waves: Germany, the Netherlands, the United Kingdom and France first, then Poland, Spain, Italy, Sweden, Switzerland and Denmark, then the remaining twenty.

The field period was 6–8 July 2026: discovery on 6–7 July, offer and site verification on 7–8 July, the search component on 8 July.

The categories describe how the 632 active providers position themselves publicly. They say nothing about delivery.

In this sample, an AEO/GEO offer is most often an addition to an existing SEO service range rather than a business built around it. Given the shared root described above, that overlap is unsurprising in itself; what the categories record is positioning, not the scope of work behind it.

Four stages were used to compare offers: analysis, strategy, technical implementation and monitoring. A stage counts only where the site describes it explicitly. For technical implementation the bar is specific — work on markup and structured data, rendering, crawler access, llms.txt or page structure. The phrase “technical SEO” alone does not qualify.

Counted separately, strategy appears in 421 records (66.6%), analysis in 382 (60.4%), monitoring in 280 (44.3%) and technical implementation in 235 (37.2%).

The 195 records describing all four stages are a subset of the 235 that describe technical implementation: 40 further providers describe implementation work without presenting the complete chain.

For 198 records no stage could be established from public pages. That is a limit of what the sources showed, not a finding that those providers perform no such work.

The practical point concerns specificity of description. The label AEO or GEO on its own states neither which tasks an engagement includes, nor who carries them out, nor which form of engine output it addresses.

The unknown share is the first thing to read here: for more than half of the active group, the presence of a price list could not be established to the evidence standard.

Among the 298 records where the status was established, 85 — 28.5% — published a figure.

These numbers describe confirmed observations. They do not establish what share of the market withholds pricing, and an unknown status does not mean a price list is absent.

Published prices also need context before they can be compared: a one-off analysis, a strategy engagement and monthly support cover different work. This study did not calculate an average market price.

The table below reports counts per country: records found, how they split between the classes, and — within the active group — how many describe all four stages, how many describe technical implementation, and what is known about pricing.

A separate pass tested two intents: finding a provider (“best GEO agency {country}”) and finding explanatory material (“what is generative engine optimization”). Each intent was run in a country wording and a Europe-wide wording, across all 30 countries — 250 queries in total, 174 provider-seeking and 76 explanatory, between 5 and 12 per country. Five control queries on IndexDock’s own domain were run and then excluded from every figure reported here: the 250 do not include them.

Those queries returned 748 unique domains. The result sets were mostly local: 570 domains — 76% — appeared in the results of a single country. At the other end, 79 domains appeared across five or more countries and 32 across ten or more; these recurring domains are predominantly English-language curated lists and agency sites with pan-European positioning.

Changing the geographic wording changed the result set substantially. The overlap was zero in 5 countries and never exceeded 26.5% in any country: “an agency for Germany” and “an agency for Europe” were, in these results, largely two different populations of websites. The medians across the 30 countries are shown below.

The study separates questions that the market usually presents as one: how a company names its service, which work it describes, whether it publishes any pricing indication, and which sites appear for which search wording.

When comparing offers, the useful questions are the scope of work, who is responsible for implementation, how monitoring is done and what the promised outcome is bounded by. The presence of the letters AEO or GEO on a website answers none of them.

The sample was assembled through search and public sources. It is not a census of European agencies and not a statistically representative estimate of the market. Coverage was uneven: from 74 records in Germany to 6 in Luxembourg, and from 5 to 12 search queries per country.

Checking depth was uneven too, which is why unknown rates differ so widely between countries and why per-country rates are not compared here.

The results describe material collected on 6–8 July 2026. Information outside the checked sources may exist. Classification of a public offer confirms nothing about service quality, client outcomes or actual visibility in AI answers.

Provider categories were assigned from public self-presentation, not from delivery. An SEO agency with an AEO/GEO offer is an established SEO, digital or marketing agency whose AEO/GEO offer sits alongside an existing service range. A specialist (pure-play) is built around AI-answer visibility, with no legacy SEO line presented as its core. An individual consultant is one named person or micro-brand with no agency structure presented. Where the type could not be determined from public material, it is recorded as unknown.