Coverage Groups

PR teams don't land one placement at a time — a healthy client earns dozens to hundreds of media hits per month. Coverage Groups let you paste that entire list into Spyglasses at once, keep it organized as a named group, and attach it to a PR project as a single goal with a single influence rate: "this month's placements account for 9% of all the sources AI cites for our tracked prompts."

Why Coverage Groups

Without them, tracking a month of placements means creating one article goal per URL — hours of manual entry, a project page listing hundreds of rows, and a client report that reads like a spreadsheet. A coverage group collapses all of that:

• One paste, one goal. Import the whole placement list in a single step. The group tracks as one goal in the project and one row in exported reports.
• One influence number. Instead of asking "did placement #37 get cited?", you get "what share of AI's sources came from our work?" — the number that justifies a retainer.
• Living lists. Add this month's placements to an existing group (or create a monthly group) at any time — attached projects pick up the changes automatically, including retroactive matching against citations that already happened.
• Every URL gets scored. Pasting a list also creates and scores a Placement for each URL automatically, so the group shows each item's AI Placement Quality Score (AIPQS) alongside the publisher's AIPVS — quality and value in one table.

Creating a Coverage Group

Navigate to Coverage Groups in the property sidebar (in the Coverage section, next to Placements), then click New group.

Give the group a recognizable name — "March 2026 placements" or "Tier-1 launch coverage" — and optionally paste URLs right in the creation dialog, one per line. You can also add URLs later with the Import URLs button on the group's detail page.

What the importer accepts

• One URL per line; comma- or tab-separated lists also work
• URLs with or without https://
• Any web URL — news articles, blog posts, YouTube videos and Shorts, LinkedIn posts, syndicated copies (MSN, Yahoo, etc.)

The importer normalizes each URL so cosmetic variants can't cause missed matches or duplicates: www. prefixes, tracking parameters (utm_*, gclid, share tokens), and trailing slashes are all unified. YouTube links get special handling — youtu.be short links, watch?v= URLs, and Shorts are expanded into companion entries so the group matches however an AI platform cites the video.

After the import you'll see exactly what happened: how many URLs were added, how many were already in the group, and any lines that couldn't be parsed (with reasons).

The Group Detail Page

Each URL in the group shows:

Because coverage items share their classification vocabulary with Citation Intelligence, your coverage mix and your citation mix are directly comparable — useful for spotting gaps like "AI cites video heavily in this category, but our coverage list has almost none."

Use Download CSV on the detail page to export the full table — every URL with its title, author, publish date, classification, AIPQS, AIPVS, and citation count — for client decks and other tools.

Scoring older groups

Groups created before placement linking existed (or whose placements were deleted) show a Score placements (N) button. One click creates and scores placements for every unlinked URL; scores fill into the AIPQS column as they complete. URLs that already have a placement are linked for free — nothing is ever re-scored.

Removing URLs

Select rows and click Remove. Removing a URL also removes the goal hits it contributed to attached projects, so influence numbers always reflect the current list — a paste mistake can't permanently inflate a report. Linked placements are kept and remain available on the Placements page.

The AI Citations Page

Click View AI citations on a group (or any Cited count) to answer the question every client asks: "when AI cites my placement, which prompt and platform did it, and what did the answer say?"

The page shows three things:

• A funnel — coverage URLs → cited by AI → cited inline (referenced in the answer text itself, a true citation, as opposed to sources the AI merely fetched while composing).
• Cited vs. not-cited comparison — how the URLs AI actually uses differ from the ones it ignores: average AIPQS, average word count, median content age, byline share, and the most-cited authors. This is the feedback loop for what kind of coverage to pitch next.
• Every citation hit — one row per citation with the coverage URL, the prompt that triggered it, the platform and model, the date, the citation type, and (for inline citations) the sentence-level context around the reference. Click a row to read the full AI answer with all platforms in tabs, opened to the one that cited you.

Filter by platform, citation type, and date range; the whole list exports to CSV. The same data is available to AI assistants through the MCP server's get_coverage_citations tool.

Attaching a Group to a Project

Two ways:

1. At project creation — PR projects show a Coverage Groups to Track checklist alongside publisher/article/author goals.
2. On a running project — use Add goal on the project's Goals card and pick the coverage group.

Either way, Spyglasses immediately checks the project's past citations against the group. If placements were already being cited, the goal shows its hits and influence rate right away and is quietly marked as Hit — no waiting for the next nightly run, and no notification blast for coverage that already existed.

From then on, every nightly prompt run records new matches. The project's goal row shows:

Coverage: March 2026 placements (46 URLs) — 14 citations, 9% of this project's citations

One group can be attached to any number of projects; each project computes its own influence rate against its own prompts.

Best Practices

Group by campaign or by month

The two natural groupings mirror how you report: a campaign group ("Product launch coverage") shows a campaign's total AI influence; monthly groups ("March 2026 placements") show how each month's work performs and make month-over-month comparisons trivial.

Import everything, including syndication

If a story ran on abc7.com and was syndicated to MSN and Yahoo, import all three URLs. AI platforms cite syndicated copies frequently, and each URL only matches exactly — the syndicated copies are separate placements earning separate citations.

Watch the Cited column before attaching

The Cited count reflects your own tracked prompts and reports, and populates as soon as data exists. A quick scan tells you which placements AI is already using — and clicking a count shows exactly which prompts and platforms — useful context before you even attach the group to a project.

Compare cited vs. not-cited to sharpen the next pitch

The AI Citations page's comparison strip tells you what distinguishes the coverage AI uses: if cited items skew toward higher AIPQS, longer articles, or fresher publish dates, that's a concrete brief for the next round of outreach.

Related

• Projects — Attach coverage groups as goals and read influence rates
• Citation Intelligence — See every source AI cites, including your placements
• Placements — The per-mention quality scores (AIPQS) shown on coverage items
• Publisher Lookup — The AIPVS scores shown on coverage items
• MCP account tools — Query coverage citations from an AI assistant with get_coverage_citations