Inside your brand kit · Audiences · Get noticed
Who should be writing about your client — found, read, and graded
Earned-coverage prospecting built into the brand kit. Map who already covers the competitors, who invites contributions, and which roundups refresh on a cycle — then have every candidate page opened and read: the byline, the verbatim evidence, the door in, and the angle that earns a place. Graded honestly, timed on a calendar, and analyzed lane by lane into the moves an agency actually runs.
Discovery · The Map
Thousands of sources in — one honest map out
Four instruments sweep the category at once. The noise gets named and set aside — what's left is ranked by how winnable it is, not by how loud it looks.
Every competitor gets its own section — arrival timelines, rising sources, and an analysis written only from the measured numbers.
Sized before you go deeper
The whole map — counts, timelines, candidates — lands from one quick discovery run.
Overlap is the signal
A page covering three of your competitors and not your client is the warmest kind of candidate.
The junk gets named
Link networks, coupon farms and paid directories are fingerprinted and sifted out — with the reason on each.
The prospecting engine, end to end
From the first discovery run to the moves your team executes — every step with the evidence attached.
Four instruments, one picture
Name up to five competitors — or none — and discovery pulls who links to them, who ranks for the category, which pages invite contributions, and who mentions brands without linking. Thousands of sources analyzed into one ranked picture, sized before anything bigger runs.
What their profiles are really made of
Each competitor gets its own section: referring domains, junk and link-network share, directory share, pages that actually rank, month-by-month arrival timelines, rising sources from the last 90 days, their newsletter and podcast circuit — and a plain-English analysis written only from the measured numbers.
Submission doors & comparison listicles
The pages that publicly accept contributions, and the roundups ranking the category right now — each lane with its own tells: doors that already cover a competitor, listicles with a year in the title, and the ones sitting in the search results buyers actually see.
Every source opened and read
Pick a lane and a depth, and an AI opens each page: who wrote it, when, which competitors it names — verbatim — whether there's a real door in, what kind of site it actually is, and the specific angle that earns a place. Rows upgrade in place from “not yet read” to a graded verdict.
Calendar opportunities
Roundups that refresh on a cycle are timed opportunities. Rows with refresh signals get a date rail — when it was last updated, and the inferred window to pitch before the next edition — so outreach becomes a calendar, not a backlog.
A lane-by-lane analysis with moves
After a read, an AI strategist reviews each lane's full set of graded rows and writes the observation, the concrete next actions — each tagged for the agency or the client, each naming a real page or person — and a deeper analysis, including a check on any grade that doesn't match its own evidence.
Refusals never count against you
If a page refuses the reader, it doesn't count against your depth — discovery keeps walking the ranked pool until your count of real reads is delivered. About a third of the web refuses automated readers; you keep those links either way.
The Read · Not Yet Read → Graded
Nothing about a link tells you if there's a way in — reading the page does
Pick a lane and a depth, and an AI opens each page: the byline, the date, which competitors it names — verbatim — the door in, and what the site really is. The row upgrades where it sits.
A candidate the map surfaced — warm on paper, unproven until opened.
“…newer platforms are closing the gap fast this year…”
updated Aug 2026 · listicle-roundup · → pitch the named editor before the next refresh
A name, not a form
Contact paths come tiered — a named writer with a channel beats a newsroom address beats a generic form. The rows say which you're looking at.
The venue gets named
A vendor's own blog, a guest-post mill, a paid marketplace — labeled for what they are, and never graded pursue without real readers.
No-fits earn their keep
A page with no way in gets read once so nobody spends an hour on it — then folds away with the reason attached.
The Strategist · Insight → Move → Delta
The analysis is where the retainer writes itself
After a read, an AI strategist reviews every lane's graded rows together and writes the observation, the moves — each tagged for the agency or the client, each naming a real page or person — and the deeper analysis, with a check on any grade that doesn't match its own evidence.
Run the whole loop from Claude
Map the category, read and grade the sources, pull the report — the same prospecting engine, driven from a conversation. Every bigger step waits for your yes.
Map the category
“Map the coverage opportunities for Acme — who's writing about our competitors and not about us?”
Read & grade the sources
“Show me Acme's coverage map — which lanes are worth reading?”
Deliver the program
“Build Acme's coverage report and give me the link.”
A refused or dead page never counts against your depth — the reader keeps walking the ranked pool until your reads are delivered.
A refresh window computed from one edition is a hint, not a proven cycle — every date rail carries the “inferred from one edition” label until more editions confirm it.
Pages that turn out to have no way in aren't hidden — they fold away with the reason each was ruled out. Knowing where not to spend an hour is part of what a read buys.
A “write for us” page on a software vendor's own blog is not a publication. The read labels vendor blogs, guest-post mills and paid marketplaces — and none of them get a pursue grade without real third-party readership.
Pages in other languages leave the read pool into a labeled fold rather than being discarded — real coverage exists in every market, but your reads shouldn't be spent there by default.
An email read off a page is a candidate, not a fact — rows say “verify before sending,” and contact resolution is its own step.