If you assess location pages, you need to understand what a city comparison can tell you about the writing. The US Cities taxonomy lets you compare a draft against a city the way other taxonomies compare it against a topic or an audience. This page explains what that comparison measures and the specific mistakes it invites.
What the comparison does
You select a city as the target category and submit your text. The report returns the ten cities whose stored positions are nearest to your writing, and tells you where your selected city falls if it is not among them.
The mechanism is the same as any other taxonomy comparison, described in how a draft is scored against a category. Your text becomes a position, and that position is compared with the stored position of each city.
Each city position is built from its hierarchical label, such as United States > Alabama > Alabaster, rather than from descriptive writing about the city. The result therefore measures similarity to the named location structure, not to a stored profile of neighborhoods, landmarks or local facts.
The taxonomy covers 4,117 US cities selected at a population threshold of 10,000 or more. There is no equivalent for other countries today, so a page written for Manchester or Toronto cannot be checked this way. Work on non-US locations has begun, but there is no release date.
What a result tells you
An unexpected category match is often the most useful finding.
Suppose a service page is written to attract customers in Boulder, Colorado. The word Boulder appears eleven times. You select Boulder as the target and the report returns Denver, Fort Collins, Colorado Springs and several other cities ahead of it.
That result says the writing reads as being about the region generally rather than about Boulder specifically. Repeating a city name is not the same as writing about a place, and this is the check that separates the two.
The phrase list alongside the cities shows terms associated with your selected city that the draft does not contain. Read it as a description of what writing about that city usually involves, which is a prompt for what you might have to say.
The mistake this measurement invites
The obvious way to raise a city score is to add local detail: neighborhoods, landmarks, nearby towns, local terminology.
That is only legitimate when the detail is true of your business. A contractor who does not work in a neighborhood should not list it. A shop with one location should not imply several. A business that has never served a suburb should not claim familiarity with it.
The measurement cannot tell the difference. It compares meaning, and invented local detail moves the number exactly as real local detail does. A page that scores well by describing work the business does not do is worse than the page it replaced, because it now misleads customers who arrive from it.
Add local details that help readers: areas served, conditions affecting the work, and applicable regulations or requirements. Answer questions customers in that city ask.
What the score is not
- It is not a local ranking signal. No search engine or AI assistant is consulted in producing it. The comparison is between your text and a stored city position. It says nothing about local search results.
- It does not replace the things local visibility depends on, such as a consistent business address, a business profile, and genuine local presence. This measurement examines writing.
- It does not verify local facts. Nothing checks that a neighborhood you named is real, that you serve it, or that a claim about local conditions is accurate.
- It does not handle multiple cities in one submission well. A page covering five service areas is about all five, and the comparison will report the nearest ones rather than confirming the one you had in mind. Check a multi-city page section by section, or accept that the result describes the blend.
Practical use
Check a page before publication rather than after. The result is most useful when there is still time to decide whether the missing local substance exists, and to write it if it does.
Where several location pages follow one template, compare two of them. Templates that differ only by city name tend to return nearly identical nearest-city lists, which is a direct measurement of the problem with that approach.
The submission limit is 8,192 bytes, so a long location page is checked in sections and the section results do not combine into one page score.
Use the score during editorial review to investigate local relevance. Check any revision against the real service area and the information a local customer needs.
See the US Cities taxonomy, optimizing local service content, or what city relevance can and cannot establish.