Use case — Competitor partner gap analysis

Find every publisher promoting your competitors — but not yet you

Your rivals' best affiliates are recruitable proof that a publisher already converts in your category. We pull your entire vertical from 120M+ classified domains, LLM-screen each site against your ICP, and flag the ones showing competitor-affiliate signals — then measure that universe against your current roster to size the gap.

~48,000
competitor-promoting publishers you don't yet partner with (estimate)
1.73M
vertical domains screened in the reference engagement
500
real gap-list domains delivered free in your pilot
The job to be done

Your competitors already validated your best prospects

Every publisher that runs a competitor's affiliate links is a proven fit: they cover your category, they monetise with affiliate partnerships, and their audience buys. The gap analysis surfaces that whole cohort, subtracts the ones you already work with, and hands you what's left.

Pre-qualified by a rival

A site promoting a competitor has already cleared audience-fit and monetisation questions. You are recruiting proven converters, not cold guesses.

Overlap you can't see

Without full-web data you can't tell which publishers your competitors depend on, or how much of that set already overlaps your roster.

A gap you can size

The value isn't a vague "we should recruit more" — it's a counted list: competitor-promoting publishers, minus your current partners, equals the recruitable gap.

How the pipeline solves it

From the whole vertical to a measured gap list

The same full-web pipeline runs, with your current roster supplied as an exclusion set so the output is only what you're missing.

1

Pull the entire vertical

We pull every classified domain in your category from the 120M+ database — the full population your competitors recruit from, not just a directory slice.

2

Screen for competitor signals

The LLM reads each site for outbound affiliate links, brand mentions, and comparison coverage that indicate a competitor relationship — each scored with confidence.

3

Apply your ICP

Competitor presence alone isn't enough. Each candidate is also screened against your required, preferred, and disqualifying signals so the list stays on-profile.

4

Subtract your roster

You supply your active partner list. We match it against the screened universe to compute overlap, then remove those domains from the recruitable set.

5

Deduplicate

Mirrors, redirects, and country variants collapse to one canonical domain, so overlap counts and gap counts are honest — never inflated by duplicates.

6

Deliver gap CSV + overlap metrics

One row per gap domain, plus a summary: how many competitor publishers you already share and how many you're missing, tuned to your ICP.

The overlap number is a diagnostic in itself. A small overlap means your program is under-penetrated in a category your rivals already work hard — a large opportunity. A large overlap means you're competitive on coverage and the remaining gap is the precise, finite list worth chasing. Either way you get a counted answer, not a hunch.
A worked example

Sizing the gap on a vertical (estimate)

A representative gap analysis. The vertical screen yields tens of thousands of publishers showing competitor-affiliate signals; matched against a roster of a few thousand active partners, most of that cohort turns out to be recruitable gap. The final figures are estimates extrapolated from calibration slices.

~52,000
publishers promoting a competitor (estimate)
~4,000
already on your roster (overlap)
~48,000
recruitable gap (estimate)
Illustrative funnel

Competitor gap, one vertical

Universe and reference counts are measured; gap and overlap figures are estimates. Your pilot returns real gap domains and a projection built from your own roster and ICP.

0
classified domains as the source
0
competitor-promoting (estimate)
0
recruitable gap (estimate)
0
delivered free in your pilot
120M+
classified domains
~52K
competitor-promoting
−4K
roster overlap
~48K
gap (est.)

Reference engagement (a large-scale production run on our own classification platform): 30M → 1.73M → 1.2M → ~96% precision on a reviewed sample. Gap figures above are estimates for illustration.

Screening signals

The specific signals behind a gap verdict

Competitor detection is signal-based and scored, never a single keyword. Each is extracted from what the site publishes and carries its own confidence.

Outbound affiliate links Competitor brand mentions Head-to-head comparisons "Alternatives to" pages Competitor coupon / deals content Relevant to your vertical Review / comparison content Fresh content (recency) Detected language Geo content focus CrUX presence & rank group Not already on your roster
Deliverable snapshot

The gap CSV, one row per publisher

An illustrative slice. Every row is a publisher promoting a competitor that you do not currently partner with; the roster-overlap rows are flagged separately so you can measure share.

A note on confidentiality

The rows below are illustrative and generalized. On public pages we withhold real domains and generalize descriptions, so any profile shown here is intentionally not traceable to a specific site — including through a web search — which protects independent publishers without changing the underlying data. Your pilot and client deliverables contain the actual domains and every signal, so everything can be verified directly.

DomainCompetitor signalICP fitConf.On your roster?LangFreshness
publisher—a1Affiliate links + comparisonFit0.93Gapen< 30 days
publisher—a2"Alternatives to" pageFit0.90Gapen< 14 days
publisher—a3Competitor coupon contentFit0.86Overlapen< 60 days
publisher—a4Brand mention + reviewsFit0.88Gapfr< 45 days
publisher—a5Affiliate links (competitor)Fit0.81Gapen< 90 days
The summary sheet is the headline. Beyond the row list, the deliverable reports total competitor-promoting publishers, roster overlap count and percentage, and the counted recruitable gap — the number that sizes your recruitment plan.
Questions

Competitor partner gap analysis — FAQ

How do you detect that a site promotes a competitor?
The LLM reads each site for outbound affiliate links, competitor brand mentions, comparison and "alternatives" pages, and competitor coupon content — each scored with a confidence value. You give us the competitor names to watch; we screen the whole vertical against them.
How do you know which publishers I already work with?
You supply your active partner list as a simple domain export. We canonicalize and match it against the screened universe to compute overlap, then remove those domains from the gap list so you only see net-new prospects.
Where does the ~48,000 gap figure come from?
It is an estimate: the competitor-promoting count extrapolated from calibration slices, minus a representative roster overlap. It is labeled as an estimate throughout. Your pilot replaces it with a projection built from your own roster, competitor set, and ICP.
Do you name the competitor a publisher promotes?
In your deliverable, yes — each gap row records which of your listed competitors the site shows signals for, so your outreach can lead with the right angle. Public pages generalize this to protect the publishers shown.
What does the free pilot include?
The full pipeline configured for your ICP and competitor set, stopped at the first 20 qualified gap domains — free — plus overlap metrics and a projected full-run gap. It is the fastest way to see who your rivals recruit that you don't.

See who promotes your competitors — but not you

The free pilot runs the full pipeline on your vertical, screens for your competitor set, subtracts your roster, and delivers your first 20 gap domains plus overlap metrics. No cost, no obligation.

Request a Free Pilot