Most directory networks die from the same wound. They scale geo pages faster than they can keep them good, and a few thousand thin city pages quietly drag the whole domain down. A fixed quality bar is the cheapest insurance against that. It is the line a page has to clear before it earns a slot in the index, and it does not move because a deadline is close.
This is the bar we hold our own city pages to across the group. It is deliberately strict, because the cost of a weak geo page is not zero. It is negative. Below I walk through the five checks, why each one exists, and how to apply them without turning your team into a bottleneck.
Why a fixed bar beats a word count
The lazy version of a quality standard is a word count. Hit 800 words and the page is approved. That rule produces exactly the content it deserves: padded paragraphs, restated headings, and a closing summary that says nothing. Search engines have been good at spotting that pattern for years, and readers spotted it long before that.
A real bar tests usefulness, not volume. A 600 word page that answers a genuine local question will outperform a 1,800 word page that could describe any city on earth. The job of the bar is to force that distinction every single time, so that no page ships on effort alone. If you want the reasoning behind why we build for depth across the whole hub, the local and geo SEO pillar sets out the wider approach.
The five checks
1. Location specificity
The page must contain facts that are true for this place and false, or simply different, for the next town over. Neighbourhood names, transport links, the local licensing quirk, the street where the good venues cluster, the months the place is busy. If every concrete noun on the page could be lifted into a different city without editing, the page is national content wearing a local costume.
This is the check most pages fail, and it is the one that matters most. We talk about it in more depth in avoiding cookie cutter city pages, because the failure mode has a name and a smell once you have seen it a few times.
2. A local answer
Every geo page exists to answer a question a person actually types or asks. Where should I stay near the old town. Which area is best for families. What does a private dining room cost here. The page has to answer that question plainly, near the top, in language a person would recognise as the answer. If a reader has to dig past three paragraphs of throat clearing, the page has failed even if the answer is buried in there somewhere.
3. A named source
Claims that carry weight carry a source. A price band, an opening season, a capacity figure, a rule about permits. Where the number comes from a public register, an operator we work with, or our own data, we say so. Where we cannot share a precise figure responsibly, we give a rounded or directional one and we say it is rounded. The point is that a careful reader can tell the difference between something we know and something we guessed.
4. A clear next step
A good geo page does not leave the reader stranded. It points somewhere useful. That might be a deeper guide, a relevant listing, or an enquiry path for someone ready to act. This is also where the page does its quiet commercial work, which we cover in geo pages that convert. A page with no next step is a dead end, and dead ends teach search engines that your pages do not satisfy intent.
5. The swap test
This is the check we are proudest of, because it is brutal and it takes ten seconds. Delete the city name from the page and read it back. If the page still makes sense for anywhere, it fails. If removing the city name makes the page read as obviously wrong, plainly about a single specific place, it passes. The swap test catches the templated filler that the other four checks sometimes let through.
How to run the bar without becoming the bottleneck
A strict bar is only useful if it does not grind production to a halt. We split the checks into two groups. The mechanical ones, broken links, missing sources, no next step, get caught by a script before a human ever looks. That frees the editor to spend their attention on the two checks a machine cannot do well: the local answer and the swap test.
We also accept that a high rejection rate is healthy. When a draft comes back for a rewrite, that is the system working, not failing. A team that passes almost everything on the first read has quietly lowered the bar without telling anyone. The honest signal is a steady stream of rewrites on the first pass and clean approvals on the second.
What the bar protects you from
The bar is not really about any single page. It is about the reputation of the whole domain. Search engines judge sites in aggregate, and a directory that publishes a thousand careful pages and two thousand thin ones is judged as a thin site with some good pages, not the other way round. Holding the line on quality is how a few hundred genuinely useful pages grow into an authority site rather than a liability.
If you are weighing whether to build this kind of disciplined geo programme yourself or partner with a group that already runs one, our building thesis explains how we think about depth, restraint, and the long game. The short version is that the quality bar is not a tax on growth. It is the thing that makes growth compound instead of decay.
The Kings Hospitality Group Geo Quality Bar: five checks, location specificity, a local answer, a named source, a clear next step, and the swap test, where you delete the city name and the page should read as obviously wrong for anywhere else.
Common questions
How many geo pages fail the bar on the first draft?
In our portfolio we expect most first drafts to need a rewrite before they pass. That is by design. A high pass rate usually means the bar is too low, not that the writing is strong.
Can the quality bar be automated?
Parts of it can. Word counts, broken links, and missing sources are easy to flag with a script. The swap test and the local answer still need a human who knows the place to read the page.