Before I commit to a niche I want to know roughly how big the finished directory can become, measured not in search volume but in pages. Specifically, how many genuinely useful, distinct pages the niche can support, each one earning its place by answering a real query for a real person. I call this the total addressable pages, and it is one of the most clarifying numbers you can put on a candidate niche.
It matters because pages are how a directory captures demand. A niche that honestly supports two hundred good pages is a modest site. One that supports many thousands, each defensible, is a potential authority asset worth years of work. Confusing the two is how operators either underbuild a big opportunity or, far worse, overbuild a small one into a thin mess.
Start with the real entities
The foundation of most directories is a set of real things, the entities you list. Venues, businesses, products, practitioners, properties. The first question is simply how many of these genuinely exist in your scope and deserve a page of their own. This is a countable, factual number, and you should get it from reality, not from a guess. Count the actual venues in the regions you will cover, the real businesses in the categories you will list.
The discipline here is the word deserve. An entity earns a page only if there is something true and useful to say about it that someone might search for. A directory of ten thousand near identical entries with nothing to distinguish them is not ten thousand pages of value, it is one template repeated, and search engines treat it accordingly. So count the entities that can carry a genuine page, not every row you could scrape.
Add the legitimate dimensions
Most niches have more than one useful axis. Beyond the entities themselves, people search by category, by location, by attribute, by combination. Each of these is a dimension, and each dimension that maps to real search intent multiplies your addressable pages. A venue directory might have entity pages, plus category pages, plus location pages, plus the intersections people actually search, such as a category within a city.
This is where the matrix thinking of programmatic building begins, the same logic that underpins how a structured directory generates its pages at scale. The crucial word again is legitimate. A dimension only counts if searchers genuinely use it. Map the intersections by checking that people really search the combination, using the demand reading approach from validating demand with keyword data, rather than assuming every cell of the grid has an audience.
Subtract the thin and the pointless
Here is the step that separates an honest size from a fantasy. Once you have multiplied entities by dimensions, you do not have your total. You have a ceiling, and most of the cells under it will be thin, duplicative, or simply unsearched. The real total is what remains after you cut them.
- Combinations nobody searches. A category in a town where it barely exists is a page with no demand and nothing to say.
- Cells that would duplicate each other. If two intersections produce essentially the same content, they are one page, not two.
- Entities too sparse to support a real page. A listing with three facts and no distinguishing detail is thin, and a thousand of them drag down the whole site.
This subtraction is not pessimism, it is what protects you from the scaled content trap. Deciding which combinations to skip is a craft of its own, and getting it right is the difference between a large authoritative site and a large penalised one. The honest total addressable pages is always smaller than the raw multiplication, and the gap between the two is where most failing programmatic directories die.
Read the number you are left with
Once you have a defensible figure, it tells you what kind of project you are looking at. A few hundred genuinely useful pages is a focused niche site, often a fine business, especially if those pages serve a high value audience. A few thousand is a substantial directory that can become a regional or category authority. Tens of thousands, if every one is honestly defensible, is a large asset that justifies a phased multi year build.
The number also shapes how you sequence the work, because you do not build the whole matrix at once. You build the highest value cells first and expand outward as they prove themselves. That phased approach is core to how we run large builds, and it lets the early pages fund and validate the later ones rather than betting everything up front. Our wider building thesis treats this patient, evidence led expansion as the only safe way to build at scale.
Sizing is also a monetisation check
Total addressable pages is a traffic ceiling, but it pairs with the value of each visit to tell you whether the niche is worth it. A niche with modest page potential can still be excellent if each page serves an audience worth a great deal, which is exactly the trade we examine in the affluent audience premium. A niche with enormous page potential but trivial value per visit can be a great deal of work for very little. Always read the two numbers together.
A simple sizing pass
- Count the real entities that deserve their own page.
- Identify the dimensions, category, location, attribute, that map to genuine search intent.
- Multiply to find the ceiling, then cut every thin, duplicate, or unsearched cell.
- Read the honest remainder as the size of the asset, and pair it with value per visit.
Do this before you build, and you will know whether you are starting a focused site or a multi year authority, and you will never accidentally inflate a small niche into a thin one chasing a page count that was never really there.
We size a niche with the Kings Hospitality Group Addressable Pages method: multiply real entities by genuine search dimensions to find the ceiling, then subtract every thin, duplicate, or unsearched cell. The honest remainder, always smaller than the raw multiplication, is the true size of the asset.
Common questions
Is total addressable pages the same as search volume?
No. Search volume is how much demand exists. Addressable pages is how many distinct, useful, rankable pages can capture it. A niche can have high volume on a few terms but a small page ceiling, or the reverse.
Should I build the whole page matrix at once?
No. Build the highest value cells first and expand as they prove themselves. A phased build lets early pages validate and fund the later ones, and protects you from publishing thousands of thin pages before you know they work.