What it costs is the biggest single topic
Share of the 261 buyer items, by the topic a page would need to cover to answer them. Topics below 1% are left out of the chart.
We collected 577 things small and mid-sized businesses said or asked about hiring an agency, ran their top questions through web search, and read 56 agency websites page by page. This is what the market looks like from the buyer's side.
Real data from public sources, 28 October 2025 to 28 October 2026. No agency is named. How it was made, and its limits
of what a business finds when it searches for "the best agency in [city]" is a directory or an agency's own list of "best agencies".
106 top-10 results, 12 searchesof what buyers raise is what it will cost, the biggest single topic. 46% of agency sites publish their own prices.
261 buyer items; 56 agency sitesof agency sites say who owns the website, domain or ad accounts, a fear buyers raise when a relationship goes wrong.
56 agency sites, 7–13 pages eachof the answers to buyers' questions come from the agencies buyers are most likely to shortlist. Publishers, platforms and a few content-heavy agencies answer instead.
498 top-10 results across the three countriesEach finding below is a pattern that held across the data, with the figure behind it and what it means for an agency.
A business looking for an agency meets intermediaries first. For "best agency in [city]" searches, 53% of the top results are directories, whose list order is partly paid, or agencies' own "best agencies" lists, and 24% of results rank their own author. Agencies' own pages are 20%.
What it means. Buyers can't take the market's shop window at face value. Proof the agency doesn't control, such as named clients, reviews in volume and referrals, counts for more than a place on a list.
What it costs is 15% of what buyers raise, the biggest topic, yet 46% of agency sites publish their own prices. AI search is 6% of the asking, and 89% of sites talk about it.
What it means. Sites are written around what agencies want to sell next, not what buyers need to decide now. A plain answer to "what will this cost me?" is the easiest gap in the market to close.
Buyers who have been burned ask how to get their domain, website or ad account back. Only 20% of agency sites say who owns them, and 32% state contract terms. Where terms are published, the detail can contradict the promise: in Australia, one site's terms charge 20–50% of the remaining contract to leave early, and another promises "no lock-in" over a six-month minimum.
What it means. Ownership and exit terms, stated up front, cost an agency nothing and answer the fear that stops a buyer committing.
The 60 agencies most visible through search, directories, reviews and awards own 6% of the answers to buyers' questions. Publishers hold 25%. Website and software platforms, which sell the do-it-yourself alternative, hold 13%.
What it means. The content that frames a buyer's decision is mostly written by others, including the agency's substitutes. Whoever answers "should I do this myself?" well shapes who gets hired.
Government and trade associations publish guidance on hiring agencies, but they are 0.8% of the answers buyers find.
What it means. There is no trusted referee in this market's answers. Whoever publishes the most even-handed answer to "how do I choose?" earns trust the directories can't.
22 of the 56 sites offer any tool, and three sites hold 46 of the 83 tools we found. Most are generic calculators, graders and AI text generators that a free chatbot can match. Only 1 builds a quote from the agency's own prices, and it asks for an email first. None gives a buyer a price for their own job without handing over contact details.
What it means. The tool buyers would use, an estimate on the agency's own rates, is the one no site offers openly. A tool keeps its value only when it needs something the agency alone has.
46% of what we collected couldn't be used, because the review sites where buyers wrote it forbid analysing their content. Reddit, the largest community where business owners talk, couldn't be reached at all. In Australia, not one buyer's post could be read in full by a permitted route, only question titles in search results.
What it means. The market's demand signal is owned by platforms. An agency's own enquiries, sales calls and client questions are a record of demand nobody else can buy.
311 usable items, coded into 109 distinct questions. Here is how the 261 items in buyers' own voice divide by topic.
Share of the 261 buyer items, by the topic a page would need to cover to answer them. Topics below 1% are left out of the chart.
Ranked by how often each was raised in buyers' own voice and by how many independent kinds of evidence agree. "Evidenced" means two kinds agree; "partly" means one. Questions are ranked, not sized: public sources give no reliable volumes.
We ran the top questions and "best agency in [city]" searches for each country and classified who owns every top-10 result.
Who owns the top 10 results for "best digital marketing, SEO and web design agency" in New York, London and Sydney, plus "how to choose a marketing agency". Results per country in brackets.
Who owns the top 10 results for each country's top 15 buyer questions (392 results). Government and associations are highlighted.
We read 56 agency websites (about 20 per country, chosen as the agencies buyers are most likely to meet) and recorded which buyer concerns each one actually answers.
Left: each topic's share of buyer asking. Right: the share of agency sites with at least one page that answers it with substance. They are two different measures, so each has its own scale.
Share of agency sites that do each. Hover or focus a bar for the split by country. "Address guarantees" includes saying that no one can guarantee rankings.
Share of the 56 agency sites offering at least one tool of each kind. Counting sites, not tools, keeps a few sites with long tool directories from dominating. Instant quotes are highlighted: the tool that would answer the most-asked question.
577 items from forums, Q&A sites, government guidance and review sites, kept only where the source's terms allow this use. 311 remained, coded into 109 questions by intent, buying stage and service. A second coder re-coded a blind 20% sample: agreement on intent and buying stage passed the 0.7 kappa threshold; agreement on who was speaking fell just short, so title-only evidence is treated as lower confidence.
498 top-10 web-search results for each country's top questions and for "best agency in [city]" searches, each classified by owner. The search tool is US-based; UK and Australian searches named the country or city, so they are not local search.
20 per country, chosen as the agencies a buyer is most likely to meet: search visibility, directory review counts (never paid list order), awards and independent roundups, with at least two signals and a real office in the country. 4 sites could not be read. The sample is provisional and no agency is named.
7–13 pages per site (8 on a typical site), read and coded against the buyer topics, plus every tool found, typed by the kind of answer it gives and whether a free chatbot could replace it.
The market report is out now, and free to read. Register your agency for its private report, which we'll email to you.