Use case — traffic-tier segmentation

Tens of thousands of qualified publishers. Which do your reps email first?

A full-web run can surface tens of thousands to hundreds of thousands of qualifying publishers in a vertical — far more than any outreach team can work at once. We tier them by reproducible traffic proxies — Chrome UX Report presence and public rank groups (top 100K / 500K / 1M) — so you sequence the highest-reach partners first. We never invent visitor numbers.

tens of thousands+
qualifying publishers a full run can surface (estimate)
4 tiers
top 100K · 500K · 1M · long tail
2 proxies
CrUX presence + public rank group
0
invented visitor counts
The job to be done

A fit verdict tells you who qualifies, not who to call Monday

Every publisher on your list clears the ICP — that's the point of screening. But a team that can run 200 outreach threads a week needs an order of operations. Tiering on reach turns a flat list of thousands into a prioritized queue.

Reach, honestly measured

Bigger publishers move more volume per partnership. We rank them using signals anyone can reproduce — not a black-box "traffic score" you'd have to take on faith.

A queue, not a pile

Reps work Tier A first for headline reach, then Tiers B and C for the long-tail volume that competitors never bother to mine.

Numbers you can defend

Every tier maps to a public, checkable data source. When a stakeholder asks "how do you know," you point at the proxy — not a made-up monthly-visitors figure.

The proxies

Two reproducible signals — and what they are not

We deliberately use only proxies that a third party could re-derive. Both are directional indicators of reach, not exact audience counts, and we label them as such.

CrUX presence

Whether a domain appears in the public Chrome UX Report dataset. Inclusion requires a site to have sufficient real-user Chrome traffic, so presence is a meaningful floor on reach — and its absence tells you a site is smaller or newer.

Public rank group

Which coarse rank bucket a domain falls into on a public popularity ranking — top 100K, top 500K, top 1M, or beyond. We use the group, never a spurious exact position, because the buckets are what the data can honestly support.

What we will not do: we don't sell you "estimated monthly visitors," "session counts," or any single hard traffic number scraped or modeled behind a curtain. Those figures vary wildly between tools and can't be reproduced. Presence-and-rank-group is coarser on purpose — and honest.
The tiers

Four outreach tiers from the two proxies

Each qualified publisher lands in exactly one tier. Tiers are a prioritization aid, not a quality judgment — a Tier D niche site can convert better than a Tier A generalist for the right advertiser.

Tier A

Top 100K

In CrUX and inside the top 100K rank group. Your highest-reach partners — often worth a personal, senior outreach touch.

Tier B

Top 500K

In CrUX, rank group 100K–500K. Substantial mid-market publishers — the workhorse of most affiliate programs.

Tier C

Top 1M

In CrUX, rank group 500K–1M. Growing niche sites — efficient to recruit at scale with templated sequences.

Tier D

Long tail

Not in CrUX or below the top-1M group. Smaller and newer — high uncontested volume, best worked with automation.

Boundaries are yours to set: the rank-group cutoffs above are the default. If your program treats "top 250K" as the priority line, or wants CrUX presence alone as the gate, the tier definitions adjust before delivery.
A note on confidentiality

The tier distribution below is drawn from real screening runs. Because the publishers are independent sites, we follow standard confidentiality practice on public pages: domains are withheld and figures generalized, so nothing here is traceable to specific sites — including through a web search. That protects the publishers without changing the underlying data. Your pilot and client deliverables carry the actual domains with each proxy value, so every tier assignment can be verified directly.

A worked example

How a qualified universe splits across tiers

An illustrative distribution for a vertical whose full run yielded roughly 60,000 qualified publishers. The shape — a small head, a broad tail — is typical; the exact split is a labeled estimate, and your own numbers come from your real run.

TierProxy definitionShare of the qualified list (estimate)Outreach play
Tier A — Top 100KCrUX + rank group ≤100K~2% · ~1,200 publishersNamed, senior, one-to-one
Tier B — Top 500KCrUX + rank group 100K–500K~11% · ~6,600 publishersPersonalized, prioritized queue
Tier C — Top 1MCrUX + rank group 500K–1M~22% · ~13,200 publishersTemplated sequences at scale
Tier D — Long tailNo CrUX / below top 1M~65% · ~39,000 publishersAutomated, high-volume nurture
Read the shape, not just the top: the head is where the headline reach lives, but roughly two-thirds of qualified capacity sits in the long tail your competitors ignore. Tiering lets you work both ends deliberately instead of stopping at the famous names.
Reproducible by design

Every tier traces to a public source

Segmentation is only useful if it's trustworthy. Both proxies are public and checkable, so a tier is an auditable claim — not a number your team has to defend on faith.

0
top rank group — Tier A cutoff
0
rank group — Tier B cutoff
0
rank group — Tier C cutoff
0
reproducible proxies, zero invented counts

Reference engagement (a large-scale production run on our own classification platform): a 30M dataset → 1.73M travel-related domains → 1.2M ICP-screened → ~96% precision on a reviewed sample. Tier figures on this page are estimates shown for illustration.

The deliverable

The tier fields in every row

Segmentation ships inside the same deduplicated CSV as the rest of the screen — sortable, filterable, and ready to drive outreach priority in your CRM.

ColumnExample valueWhat it does for you
crux_presenttrue / falseThe reach floor — is the site in the public CrUX dataset
rank_grouptop_500kCoarse public popularity bucket — never a fake exact rank
traffic_tierBThe single A–D label your reps sort and sequence on
proxy_sourceCrUX + public rankingThe auditable origin of the tier, for stakeholder defense
is_icp + confidencefit · 0.9Tiering only ever ranks publishers that already qualify
Note: tiers rank an already-qualified list. A domain is never included because of its traffic — only ordered by it. Fit is decided first; reach only sets the sequence.
Questions

Traffic-tier segmentation — FAQ

Why not just give me monthly visitor estimates?
Because they're not reproducible. Third-party "monthly visitors" figures swing wildly between tools and can't be re-derived or audited. CrUX presence and public rank groups are checkable by anyone, so a tier is a claim you can stand behind — which matters more for prioritization than a precise-looking number that's quietly a guess.
What does CrUX presence actually tell me?
The public Chrome UX Report only includes origins with enough real-user Chrome traffic to report on. So presence is a meaningful floor — the site clears a real-usage threshold — and absence flags a smaller or newer publisher. It's directional, not an exact audience size, and we label it that way.
Can I change the tier boundaries?
Yes. The top 100K / 500K / 1M cutoffs are defaults. If your team prioritizes on a different line — say top 250K, or CrUX presence alone — we set the tier definitions to match before delivery, and the underlying proxy values stay in the file so you can re-cut later.
Should I ignore Tier D?
No — the long tail is usually the majority of qualified capacity and the part competitors never work. Tier D is best handled with automation rather than one-to-one outreach, but for many advertisers a niche Tier D site converts above a generalist Tier A one. Tiers set sequence, not worth.
Does traffic ever affect whether a publisher qualifies?
No. Fit is decided by your ICP signal set first; tiering only orders the publishers that already passed. A domain is never added to your list because it has traffic — only ranked once it qualifies on the merits.
Can I see this before committing?
Yes. The free samples show the output format, and the free pilot runs the full pipeline on your vertical and returns your first 20 qualified publishers — already tiered — plus a full-run projection.

Give your reps a prioritized queue, not a flat list

The free pilot runs the full pipeline on your vertical, tiers the results by reproducible traffic proxies, and delivers your first 20 qualified publishers plus a full-run projection. No cost, no obligation.

Request a Free Pilot