Guide — vertical sizing

How to measure the TAM of an affiliate vertical

Most advertisers wildly underestimate how many publishers exist in their space, because they only count the ones they can already see. Sized from a 120M+ classified-domain base, a single vertical typically holds tens of thousands to hundreds of thousands of qualifying publishers — and multi-country scope pushes the candidate pool into the millions.

Hundreds of thousands
qualifying publishers a large vertical can hold (estimate)
Millions
candidate domains once scope spans multiple countries
120M+
classified domains the sizing is drawn from
The undercount

Why your vertical looks smaller than it is

Publisher TAM is not "how many affiliates are in my network." That figure is a rounding error against the real market, because it only counts the sites that opted in. Sizing from a classified index of the whole web tells you the true addressable field — most of it uncontested.

You count what you can see

Network dashboards, past outreach, and page-one search results are a tiny, visible slice. The long tail sizing lives outside all three.

Geography multiplies it

The same vertical exists in every language and market. A US-only view can miss the majority of the worldwide candidate pool.

Formats multiply it again

Blogs, comparison hubs, newsletters, deal sites, video-led creators — each format adds a layer of publishers a single-format search never surfaces.

Magnitude framing

Think in orders of magnitude, not dozens

Before any precise number, calibrate your intuition to the right scale. These bands describe how publisher TAM tends to distribute — narrow to broad, single-market to global. Every figure here is an estimate for framing, not a guarantee.

Narrow sub-niche
Thousands

A tightly-scoped single topic in one country and language.

Full vertical
Tens of thousands

An entire category, one primary market, after ICP screening.

Broad vertical
Hundreds of thousands

A large category with many formats and adjacent sub-niches.

Multi-country
Millions

The same vertical across many languages and markets — candidate domains before ICP screening.

Say it in magnitudes: when you brief a stakeholder, "tens of thousands to hundreds of thousands of qualifying publishers" is both truer and more useful than a false-precision single number. The exact figure comes from a run on your ICP.
The method

Sizing a vertical in four steps

TAM is not guessed — it is funneled down from a known starting universe, with each narrowing labeled and inspectable.

1

Anchor the universe

Start from the 120M+ classified-domain base — the whole indexed web, not a network roster or a search sample.

2

Pull the category

Select every domain the classifier associates with your vertical. This is the raw category count, before any quality screen.

3

Apply the ICP rate

From a reviewed calibration slice, measure the share that clears your ICP. Multiply through to estimate ICP-fit publishers.

4

Label the estimate

Report the result as an estimate with its basis stated. Scope by country or format to widen or narrow the number.

Worked funnel

From category domains to ICP-fit publishers

An illustrative sizing for a mid-large consumer vertical, one primary market. The category and ICP-fit counts are labeled estimates from the classification base; a real run on your ICP replaces them with measured numbers.

0
classified domains (the source)
0
category domains (estimate)
0
ICP-fit publishers (estimate)
0
delivered free in your pilot
120M+
classified domains
~900K
category (est.)
~72K
ICP-fit (est.)
500
free pilot slice

Reference engagement (a large-scale production run on our own classification platform): 30M dataset → 1.73M travel-related domains → 1.2M ICP-screened → ~96% precision on a reviewed sample. The funnel above is illustrative and estimated.

Publisher TAM

The count of distinct, ICP-fit publisher domains that exist across the whole web for a given vertical and scope — deduplicated to one canonical domain per site. It is an addressable-market figure, not a promise of reachable or responsive partners.

Reading the number honestly

What the estimate is — and isn't

A sizing figure is a labeled extrapolation, useful for planning scale and budget. It is not a headcount of partners who will reply.

It is

An estimate of addressable ICP-fit domains, funneled from a known universe with a stated basis and scope.

It isn't

A guaranteed count, a response-rate forecast, or an invented traffic number. Precision comes from a run, not a formula.

Watch-outs

Four mistakes that shrink the number on paper

Most sizing errors are one-directional: they make a vertical look smaller than it is, and the addressable field quietly disappears from the plan. Each of these is easy to correct once named.

Counting the network only

Treating your affiliate network's roster as the market. It is the enrolled slice, not the vertical — often a small fraction of the addressable field.

One-market blinders

Sizing only the home market and ignoring the same vertical in other languages, where much of the candidate pool actually sits.

Single-format thinking

Counting blogs but not comparison hubs, newsletters, deal sites, or video-led creators — each a distinct layer of publishers.

Search-result sampling

Extrapolating from page-one results, which surface the visible few and hide the long tail that never ranks yet still fits.

The correction: anchor to the whole classified universe, scope by country and format on purpose, and let a run — not a search box — set the number. Each fix moves the estimate up, toward the real magnitude.
Questions

Sizing an affiliate vertical — FAQ

Where do the universe and category numbers come from?
The 120M+ figure is the size of our classified-domain database. Category counts are the domains the classifier associates with a given vertical. The ICP-fit figure is an estimate, extrapolated from a reviewed calibration slice and always labeled as an estimate.
Why give a magnitude range instead of one exact number?
Because the honest answer before a run is a range. Verticals hold tens of thousands to hundreds of thousands of qualifying publishers depending on scope, and a single false-precise figure would imply certainty we don't have until your ICP has actually run against the category.
How does multi-country scope change the number?
Substantially. The same vertical exists in every market and language, so spanning multiple countries can push the candidate pool into the millions of domains before screening. Detected language and geo focus let you scope the sizing to exactly the markets you sell into.
Does TAM mean I can reach all of them?
No. TAM is the addressable field of ICP-fit domains, not a forecast of how many will respond or convert. It sizes the opportunity and sets the ceiling for a recruitment program; response rates depend on your outreach, offer, and prioritization.
How do I get a real number for my vertical?
Run the free pilot. It configures the full pipeline for your ICP, returns the first 20 qualified publishers, and includes a full-run yield projection specific to your category and scope — replacing the illustrative estimates with a measured basis.

Get a real yield projection for your vertical

The free pilot runs the full pipeline on your category, delivers your first 20 qualified publishers, and projects the complete-run yield for your exact ICP. No cost, no obligation.

Request a Free Pilot