When a local business ranks, does it own the page?

Measured 2026-08-11 to 2026-08-11. Published by JustMarket.Me. Data free to reuse with attribution (CC BY 4.0).

Across 90 local service searches in 9 American towns, a business’s own website held 21.3% of top-three organic results in the smallest towns against 38.6% in the largest metros — the opposite of the direction predicted before measuring — while platforms and directories took 67.4% of the small-town results.

What was measured, and how

On 2026-08-11 we ran 90 live searches. The frame is 30 consumer-facing local verticals — restaurants, salons, trades, clinics, garages, groomers, photographers and the rest — spread across 9 sectors and written out in full before any search ran. Each vertical was searched in three towns, one from each of three city-size tiers, using a fixed assignment rule so that no vertical is confined to a single town and each of the 9 towns carries the same number of verticals. The query was identical every time: the vertical, the word "in", the town and the state.

For each query we recorded the first three organic results — title and URL, verbatim — giving 270 result slots across 137 distinct hosts. Every slot was then classified on two axes defined in advance and kept deliberately independent. The first asks how many businesses the page is about: one specific business, many businesses, or no particular business. The second asks who controls the domain it sits on: the business itself, a platform whose trade is listing other people’s businesses, a publisher, or an institution.

Keeping those two axes independent is the whole design, because the headline is a cross-tabulation of them. If the domain were allowed to decide the subject — assuming that a platform page must be a list, or that a business’s own domain must be about only itself — the study would have manufactured the correlation it claims to have found. Both rules therefore read only the title string and the URL path, every row records the reason it was classified as it was, and where the rules genuinely do not decide, the slot is recorded as low confidence and dropped from that axis rather than guessed at.

Six numbered predictions were written into the protocol before measuring, and they are scored on this page rather than quietly forgotten: 3 of 6 held. So were 5 conditions that would have stopped publication — a minimum sample, a minimum in each tier, a ceiling on unclassifiable slots, a floor under the number of single-business results the headline rests on, and a requirement that the sector structure not come out flat. All 5 passed. One amendment to the classification rules was made after the first batch, before the bulk of the data existed; it is written into the protocol with the case that forced it.

What this does not tell you

• No map pack, no ads, no AI overview, no personalisation. This describes the ordinary organic links and nothing above them. On local searches a great deal sits above them, and the map pack in particular is where much of this traffic actually goes. It is invisible to this measurement, and a business absent from these results may be doing perfectly well in the pack.
• Titles and URLs only. No page was fetched. Whether a ranking page is any good, whether the profile is complete, whether the site converts anybody — none of that is visible here and none of it is claimed.
• "Small town" means three named towns, not a sample of American towns. This is the limit that constrains the headline hardest, so it is worth being blunt: the tiers are three towns each, not a sample of American towns, and the tier labels are a coarse label chosen in advance, not a population cut-off. No population figure was measured and none is stated. City size is confounded with the particular towns picked, and a different nine towns could move the gap. What the study can say is that across these towns the direction was consistent and it was the opposite of what was predicted.
• One reading per query on 2026-08-11, from one US-locale backend, in English. Results move. A single reading cannot tell a durable pattern from the state of one afternoon.
• 30 verticals, enumerated, not sampled. There is no register of local verticals to sample from, so the frame was written by hand before measuring. The per-sector and per-tier comparisons carry the weight; an average over a hand-written list carries less, and several sector cells are small. The counts are printed beside every percentage so a reader can see how thin each is.
• Who controls a domain is a judgement from the domain and the title. A booking platform that also hosts a business’s only web presence, a manufacturer running a contractor locator, an insurer running a provider directory — these sit between categories. Fixed rules were written in advance and applied consistently, but a reader could reasonably draw the line elsewhere.
• Owning the page is not the same as being found, and this study says nothing about whether owning the page is better. It measures which kind of page occupied the slot. Whether a business gets more work from its own site than from a platform profile is a question about customers, not about search results, and nothing here answers it.
• 3 of 270 slots could not be classified confidently on domain control and are excluded from every figure on that axis. The unresolved cases are all the same shape: a branded domain that could be a single operator or a marketplace, which cannot be settled from a name and a title.
• This is one reading of what ranked, not why it ranked. Nothing here is about search volume, traffic or ranking difficulty; there is no keyword tool in this measurement and no such figure appears in it.

The smaller the town, the less of the page a local business owns

This is the question the study was built to answer, and the prediction was written down in advance because the intuition is so strong: a small town has fewer businesses competing, so a local business’s own website should find a clearer field. We committed to an own-domain share at least eight points higher in small towns than in large metros. The gap came out at -17.3 points — the other way round.

In the largest metros, 38.6% of top-three organic slots sat on a business’s own domain. In mid-size cities it was 38.9%, essentially unchanged. In the smallest towns it fell to 21.3%, while platforms and directories rose from 51.1% of metro slots to 67.4% of small-town ones.

The sharpest single number is the renting one. A page about one specific business, hosted on somebody else’s platform — a Yelp business page, a Facebook page, a booking profile — was 3.4% of slots in the large metros and 12.4% in the small towns. The share of results that were about one business at all barely moved between tiers, at 35.6% and 35.6%. What changed was who owned the page that business appeared on.

Why is beyond this instrument, and the limits above say so. One reading of what ranked cannot say what caused it. What can be said is that the easy-field intuition, in these towns on this day, was backwards.

Most of the page is not a business at all — but when it is, the business usually owns it

Before asking who owns the ranking page, it is worth asking how often the ranking page is about a business in the first place. Across every slot classified confidently, 60% were pages presenting many businesses at once — a directory’s category page, a platform’s search results, a round-up article — and 3.3% were about no particular business. Only 36.7% were about one specific business.

Within that 36.7%, the answer is the reassuring one, and it was predicted correctly. Of the 97 results about a single named business, 75.3% sat on that business’s own domain and only 19.6% on a platform profile. When a local business does occupy a slot in its own right, it is usually occupying something it controls.

Put the two together and you get the number that actually matters to somebody deciding a budget. A slot that is both about one business and on that business’s own domain — the thing a website can win — was 27.3% of all measured slots. In 38 of the 90 searches, not one of the three results was a business’s own page. In 23 searches every one of the three sat on a platform.

Sector decides more than city size does

The tier gap is the surprise, but it is not the largest effect in the data. Own-domain share ran from 66.7% in auto down to 16.7% in food and drink — a spread of 50 points, against -17.3 across the city-size tiers.

We predicted that home trades would be more platform-dominated than health, by at least ten points, on the reasoning that Angi, Thumbtack and their peers have made trades their business. That held: health came in at 50% own-domain against 35.6% for home trades, a gap of 14.4 points. What was not predicted is how far below both of them the consumer-facing sectors sat.

Two of the 30 verticals returned no own-domain result at all across any of their three towns: restaurant and pet boarding. We had predicted at least five would, and were wrong — most verticals manage at least one. The cells here are small, and the counts are printed beside each percentage precisely so that nobody reads a sector line as more solid than it is.

What we predicted, and where we were wrong

Six predictions were committed before the first search. 3 held and 3 did not, including the one the study was built around. They are printed here because a study that only reports the predictions it got right is not reporting a method, it is reporting a conclusion.

The two misses beyond the tier direction are worth naming. Platform-hosted profiles of a single business turned out to be 7.1% of all slots, below the ten per cent floor we set — the platforms dominate through their directory and category pages, not through individual business profiles. And 2 verticals returned no own-domain result at all, against the five or more we expected.

Two things we did not plan to measure

Neither of these was in the protocol, so both are labelled and neither is the headline. A pattern noticed after seeing the data is weaker evidence than one committed to beforehand, however clean it looks.

The first is concentration. 137 distinct hosts appear across 270 slots, but a single one — yelp.com — accounts for 26.7% of every slot measured, appearing 72 times. Nothing else comes close: the next host, wikipedia.org, holds 2.2%. One company is a quarter of the organic local results in this sample.

The second concerns what "own domain" conceals. Of the 73 slots where a single business held its own page, 40 were what a reader would call an independent local business and 33 were a chain or franchise location — a national brand’s page for that town. The classification scheme was written in advance to count a franchise location on its corporate domain as own-domain, which is defensible, but it means "the business owns its ranking page" and "an independent local business owns its ranking page" are not the same statement, and this study measured the first.

Checking this yourself

Pick your trade and three towns of different sizes you actually serve. Search the trade, the word "in", and each town, and ask two questions of each of the first three results: is this page about one business or many, and is the domain that business’s own. That is the entire method, and a handful of searches is enough to see which end of this range your own market sits at.

The protocol was committed before the first search, with the vertical list, the town list, the assignment rule, both classification schemes, the publication minimums and the 6 predictions written down in advance. The instrument probe that fell outside the frame was discarded and logged rather than kept. The raw rows are committed with the reason for every classification, and the script that turns those rows into every figure on this page is committed beside them. Re-run it and you get these numbers, or you get a bug report worth having.

Or start from ours. Every figure on this page is published as JSON at /functions/serveResearchData, under CC BY 4.0 — free to chart, re-cut or quote, with attribution, and with the method and limits travelling alongside the numbers. New studies arrive by RSS.