Use case — Publisher recruitment at scale

Recruit net-new publishers across a whole vertical — not the same marketplace list everyone shares

Marketplaces and networks hand every advertiser the same enrolled roster, so your program competes for the identical few thousand names. We start from 120M+ classified domains, pull your entire vertical, and LLM-screen each site against your partner ICP — surfacing tens of thousands to hundreds of thousands of qualifying publishers your competitors never see.

~122,000
qualifying travel publishers from a full run (estimate)
1.73M
travel-related domains identified (reference engagement)
500
real qualified domains delivered free in your pilot
The job to be done

Your recruitment ceiling is the network's roster, not the market

Most programs plateau because the recruitable universe they can actually see is tiny: the publishers already enrolled in the same two or three networks. The real vertical is orders of magnitude larger, and almost all of it has never been contacted. The problem is discovery, not outreach capacity.

Everyone recruits the same names

The enrolled marketplace list is public to every competing advertiser. Bidding for the same publishers pushes commissions up and differentiation down.

The long tail is invisible

Single-operator niche sites, regional blogs, and category authorities that never joined a network simply don't appear in directory-based tools.

Manual sourcing doesn't scale

A team googling "best travel blogs" reviews a few hundred sites a quarter. A vertical holds tens of thousands to hundreds of thousands. Hand-search never closes that gap.

How the pipeline solves it

From the whole web to a ranked recruitment list, in six steps

The same repeatable pipeline runs on any vertical. Each step is explicit and inspectable — you see the universe size, the ICP definition, the calibration sample, and the confidence on every field.

1

Pull the entire vertical

Your industry maps to categories in our 120M+ classified-domain database. We pull every active domain in scope, including the long tail no directory has profiled.

2

Define your ICP as signals

Your ideal publisher becomes a written set of required, preferred, and disqualifying signals — not a fuzzy prose brief and not a keyword list.

3

Calibrate on a reviewed sample

We screen a slice, your team marks fit vs not-fit, and we tune the signal definitions until measured precision clears your bar. Only then does the full run start.

4

Full-run LLM screening

An LLM reads every domain in the universe and extracts each signal with its own confidence score. ICP fit is computed from the signal set, never guessed.

5

Deduplicate to one row per domain

Mirrors, redirects, and country variants collapse to a single canonical domain, so outreach never contacts the same publisher twice under different names.

6

Deliver CSV + full-run projection

A structured CSV, one row per canonical domain, ready for your partner CRM — plus a projected yield for the complete vertical, tuned to your ICP.

Nothing here is a black box. You see the universe size before screening, the exact required, preferred, and disqualifying signals in your ICP, the reviewed calibration sample, and a confidence score on every extracted field. The output scales from a single sub-niche to an entire multi-country vertical without changing the method — only the scope dial moves.
A worked example — travel vertical

~122,000 qualifying publishers from one full run (estimate)

Here is how the numbers scale on a real vertical. In the reference engagement — a large-scale production run on our own classification platform — the pipeline identified 1.73M travel-related domains and ICP-screened 1.2M in a single run at ~96% precision on a reviewed sample. Applying a representative independent-content ICP to that vertical projects the qualifying-publisher count below.

Illustrative full run

Travel vertical, independent-content ICP

The final figure is an estimate extrapolated from calibration slices; your pilot replaces it with a projection built from your own ICP and a real 20-publisher sample.

0
classified domains as the source
0
travel-related domains identified
0
qualifying publishers (estimate)
0
delivered free in your pilot
120M+
classified domains
1.73M
travel-related
~122K
qualifying (est.)
500
free pilot slice

Reference engagement figures (30M → 1.73M → 1.2M → ~96%) are measured; the ~122,000 qualifying-publisher figure is an estimate shown for illustration.

Read the magnitude, not just the sample. A single content vertical yields tens of thousands to hundreds of thousands of qualifying publishers. The 20-publisher pilot slice is the measured basis for your projection, not the size of the opportunity.
Screening signals

Specific signals, scored per domain

Recruitment quality comes from screening on signals a partnerships lead would actually check — each extracted from what the site publishes and shipped with a confidence score, so you can prioritise the highest-certainty publishers first.

Relevant to your vertical Blog-like editorial structure Individual or small team run Affiliate links present Display ads present Newsletter Active social presence Fresh content (recency) Evergreen guides Review / comparison content Detected language CrUX presence & rank group Own commerce (disqualifier) Thin AI-spun content (disqualifier)
Deliverable snapshot

One deduplicated CSV row per publisher

An illustrative slice of the recruitment CSV. Domains are withheld here for confidentiality; your pilot and client deliverables carry the real canonical domains alongside every field and confidence score.

DomainTypeICP fitConf.MonetizationLangFreshnessRun by
publisher—01Destination blogFit0.94Affiliateen< 30 daysIndividual
publisher—02Flight-deals siteFit0.91Affiliate + adsen< 14 daysSmall team
publisher—03Itinerary guidesFit0.88Affiliateen< 60 daysIndividual
publisher—04OTA (sells inventory)Excluded0.97Own commerceen< 7 daysCompany
publisher—05Regional travel magFit0.83Adsde< 90 daysSmall team
Every column is a screening field, not a scrape. The full CSV adds geo focus, newsletter and social presence, evergreen-vs-news mix, and a per-signal confidence score — importable straight into your partner CRM or outreach stack.
Questions

Publisher recruitment at scale — FAQ

How is this different from an affiliate network's publisher directory?
A network only shows publishers who already enrolled with it — the same roster every competing advertiser recruits from. We start from the whole classified web, so the output includes the long tail of independent sites that never joined any network.
Where does the ~122,000 travel figure come from?
It is an estimate extrapolated from calibration slices applied to the 1.73M travel-related domains identified in our reference engagement. It is labeled as an estimate. Your pilot replaces it with a projection built from your own ICP and a real 20-publisher sample.
Can you narrow recruitment to a sub-niche or region?
Yes. Scope is a dial. We can constrain the pull to a sub-vertical, a detected language, or a geographic content focus, and the run returns only publishers matching those constraints plus your ICP.
How do you keep competitors and advertisers out of the list?
The own-commerce signal is a disqualifier by default, so sites that sell inventory themselves are removed structurally rather than by hand. You can adjust the disqualifying set during calibration if you want to keep a specific class of publisher.
What does the free pilot include?
The full pipeline configured for your ICP, stopped at the first 20 qualified publishers — free — plus a projected full-run yield and quotes. It is the fastest way to see real names in your vertical before committing to anything.

Recruit the publishers your competitors can't see

The free pilot runs the full pipeline on your vertical, tuned to your ICP, and delivers your first 20 qualified publishers plus a full-run projection. No cost, no obligation.

Request a Free Pilot