Outreach
The publishers that cite your competitors but not you, ranked by what the gap costs, each with a researched pitch you can send.
On this page
Citations shows you where you are missing. Outreach is the tab that does something about it: the publishers that cite your tracked competitors on pages where your brand never appears, ranked by how much tracked-prompt visibility each gap costs you.
Open it from Citations → Outreach. The /outreach route redirects there. It is hidden in the white-label client portal, which is read-only.
How the queue is built
Nothing here is a wishlist. Trakkr computes the list live from the citation snapshot it already collects, so an opportunity only exists when real AI answers cited a page that names a rival and not you.
Domains where outreach is impossible or inappropriate are excluded outright: .gov, .edu, .mil, academic sources, and Wikipedia. A second group is kept but pushed down the ranking, because there is no editorial contact path to pitch: Reddit, Quora, YouTube, the large social networks, Amazon, Yelp, Trustpilot and Medium. Pages owned by a competitor are down-ranked for the same reason.
If the queue is empty, Trakkr found no publisher citing a tracked competitor where you are absent. That is the good outcome, not a bug.
Impact, and what a quick win means
Each publisher carries an impact score from 0 to 100. It answers one question: how much tracked-prompt visibility is this publisher costing you? The score rises with how often engines pulled these pages, how many of your tracked prompts and queries they answer, how many rivals are cited on them, and the domain's authority. It falls when the domain cannot realistically be pitched.
Impact is damage, not opportunity. A competitor's own homepage leaks a lot of visibility and can never be won, which is why the penalties exist.
Trakkr also estimates a difficulty of Low, Medium or High, relative to your own set of gaps rather than an absolute scale, using appearance counts and Domain Rating. A publisher that is both top-ranked and plausibly easy gets a Quick win flag. It is deliberately rare.
Every row shows the domain, its source type, how many competitors are cited there, its impact, the number of gap pages and prompts affected, and its domain rating.
The three lenses
The list on the left switches between three ways of cutting the same gap set. The right pane follows whichever one you are in.
| Lens | What the list holds | What the right pane gives you |
|---|---|---|
| Publisher | The relationships to win, ranked by impact | The full publisher plan: brief, pitch, evidence |
| Prompt | The tracked prompts you are most absent from | A "win this prompt" read, with the publishers behind it |
| Competitor | Who is cited where you are not | Their battlegrounds, and where to attack |
Clicking a battleground from the Prompt or Competitor lens switches you to the Publisher lens on that domain, so the list and the pane stay in step. In the Publisher lens, status pills filter to All, New, Contacted or Won.
Reading a publisher
The right pane fills in three weights.
The brief. The domain, its impact score, a winnability gauge, its domain rating, and one plain sentence carrying the whole gap: how many rivals are cited on how many pages here, how many of your tracked prompts those pages answer, and how often they were cited in total. Every number is hoverable and defined. Ask Agent opens the Agent with this publisher in context.
The pitch. Trakkr reads the actual page before writing anything. A research agent fetches the gap page and its author, about and contact pages, then returns a dossier: whether the gap is still live, what the page's inclusion bar is, how entries get tested, who to contact, and any time-sensitive detail the snapshot cannot know. That dossier is keyed by page rather than by brand, so research any customer has already paid for makes the next brand's answer instant.
The draft is written from that dossier plus a bounded sheet of facts about your brand. Two rules are enforced rather than requested: every claim about your brand has to trace back to the fact sheet, and the proof quote has to be a character-for-character substring of the page text. A draft whose quote fails that check does not ship as verified.
The folds. Collapsed by default, opened when you need them:
- The exact on-page edit, a paste-ready entry in the page's own format, where it slots in, and the wedge that justifies it
- How they pick, the inclusion bar and how fresh the page is
- Why you're out, the gap pages themselves with the rivals and prompts on each
Sending, and what happens after
Connect Gmail and Send posts the pitch from your own mailbox, so it lands in your Sent folder and threads normally. Trakkr only sends; it does not read your inbox. If you would rather compose yourself, Open in hands the draft to Gmail, Outlook, Apple Mail, or your clipboard. Redraft regenerates it.
A publisher moves through four states: New, Contacted, Won, plus Not a fit and Snooze to clear it out of the way. Marking it Contacted starts the watch.
You do not have to decide whether you won. Trakkr snapshots how many times the publisher mentioned you at the moment you made contact, then watches the citation snapshot and flips the row to Won the first time you appear more often than that. The banner names the date, how many days it took, the pages you are newly cited on, any prompts you recovered, and a link to where you appear. While you are still waiting, the pane says so plainly, and shows whether rivals are on fewer pages here than when you reached out.
This is observational. Trakkr reports that you appeared after you reached out; it does not claim your email caused it.
Citation gaps in Actions
Outreach is one of two paths. Citation gaps also arrive as Suggested actions with the source page, its evidence and a work type attached, and those carry a frozen measurement plan.
An outreach draft gets a 42-day window that watches citations from the target source and whether that source starts mentioning your brand. When the window closes the work lands on Earned, No movement, Coincided with a drop, or Couldn't measure, and the reasoning is in Results. Starting outreach work is also one of the few reasons Trakkr creates an external page record, so the target URL keeps its citation evidence, linked work and history in one place.
Common questions
My queue is empty. Why?
Outreach is computed from citation gaps, so you need at least one citation report to have landed, which takes a day after your first prompt run. After that, an empty queue means Trakkr found no pitchable publisher citing a tracked competitor where you are absent. If your prompt set is narrow, widen it in Research and more gaps will surface.
Can I add a publisher by hand?
No. The queue is built only from domains AI actually cited for your prompts, which is what stops it filling with sites nobody reads. If you have a domain in mind, check Citations → Sources first. If AI is not citing anything on it across your category, the leverage is lower than it looks.
Why is a big-name site ranked below a smaller one?
Because impact measures the gap, not the brand. A domain with no editorial contact path, or one owned by a competitor, is penalized however much visibility it holds. A niche publication that answers four of your tracked prompts and lists three rivals will beat a famous site that answers one.
Do I have to mark something Won?
Only if you want to. Trakkr flips the status itself when it sees you newly cited on the publisher, comparing against the mention count captured when you marked it Contacted. Marking Won by hand just moves it out of the open queue sooner.
Is the pitch safe to send as written?
Read it first. The claims about your brand are constrained to a verified fact sheet and the proof quote is checked verbatim against the page, but the angle is a starting point. You know your positioning better than the model does, and anything the draft marks for you to confirm needs confirming.
Should I run this in-house or hand it to someone?
Either works. The dossier, the page-level evidence and the exact edit are designed so a contractor can pitch without learning your category first, which removes most of the briefing. Teams that keep it in-house usually do so because editor relationships compound, and the second pitch is much easier than the first.