--- title: "How to Audit a PR Agency Coverage Report, Line by Line" description: "A four-bucket method and scoring sheet to sort any PR coverage report into earned, syndicated, sponsored and owned, so you see what your retainer really bought." author: "Shilika Jain" date: "2026-10-28T01:39:50.255+00:00" tags: ["pr measurement", "pr agency", "earned media", "crypto pr", "web3 pr"] canonical: "https://www.shilikajain.com/blog/pr-report-audit-paid-links-vs-earned-coverage" --- # How to Audit a PR Agency Coverage Report, Line by Line By [Shilika Jain](https://www.shilikajain.com/authors/shilika-jain) - 10/28/2026 A four-bucket method and scoring sheet to sort any PR coverage report into earned, syndicated, sponsored and owned, so you see what your retainer really bought. --- # How to Audit a PR Agency Coverage Report, Line by Line To audit a PR agency coverage report, sort every line into one of four buckets: earned (a journalist chose to write it), syndicated (your release copied onto partner sites), sponsored (you paid for the slot) and owned (your own channels). Then score only the earned lines for quality, and compare that score with what you're paying. Most reports look very different once the 200 links are split into four columns and only one column is counted as coverage. I've written reports for clients and I've read a lot of other agencies' reports when founders ask me for a second opinion. The method below is the one I use for both. ## Why a coverage report needs auditing at all A coverage report is a document the agency writes about its own performance. That's not a reason to distrust it, but it is a reason to check it. The common problem isn't fabrication. It's mixing: wire pickups, paid articles and genuine stories all listed together under "coverage", with one total at the top. Founders also tend to audit the wrong party. I noticed this with a prospective client earlier this year. > companies hire agencies w/o asking too many questions or hard bounding them on the KPIs because those agencies are famous, but when they don't deliver they hardbound freelancers or small contractors for all the things that they agency couldn't achieve!! > > [Shilika Jain on X, 1 Jun 2026](https://x.com/Shilika_jain/status/2061476232034996344) Hold everyone to the same standard, and start with the report you already have. ## The four buckets, and how to spot each one Open the report in a spreadsheet. Add a column called "bucket". Go line by line, open every URL, and use these tells. | Bucket | Definition | Tells on the page | | --- | --- | --- | | Earned | A reporter or editor chose to publish it, in their words | Staff or contributor byline, your news mixed with context or other sources, quotes that came from an interview, no paid label | | Syndicated | Your release republished as-is on partner sites | Byline reads "Press Release" or the wire's name, identical text on many domains, URL paths like /press-releases/ or /pr/, same timestamp across sites | | Sponsored | You or the agency paid for the slot | "Sponsored", "Partner content", "Paid post" or "Advertorial" labels, disclaimer at the top or bottom, separate sponsored section in the site menu | | Owned | Your own channels | Your blog, your X account, your newsletter, your Medium, your founder's LinkedIn | A few edge cases come up in almost every audit. - **Press release sections on real outlets.** Several big crypto sites run a press release section alongside their newsroom. A line can carry a famous outlet's name and still be syndicated. Check the byline and URL path, not the logo. - **Contributor or council posts.** Some outlets sell membership programmes that let you publish under their brand. If a fee was involved, it's sponsored. - **Podcasts and AMAs.** A podcast that invited your founder on editorial grounds is earned. A slot bought as part of a package is sponsored. Ask. - **Aggregators.** News apps and price sites that auto-ingest feeds are syndicated, even if the headline looks native. ## The scoring sheet Once every line has a bucket, only earned lines get a quality score. The other three get a zero for coverage value, though you still record them because they tell you where money and time went. Score each earned line out of 10. | Factor | 0 points | 1 point | 2 points | | --- | --- | --- | --- | | Outlet fit | Outlet your buyers and investors don't read | Relevant trade or regional outlet | Tier-1 for your category (for crypto: CoinDesk, Cointelegraph, The Block, Decrypt, Blockworks; or a mainstream business outlet) | | Depth | One-line mention in a roundup | Paragraph with context | Story about you, or a feature with your founder quoted at length | | Message pull-through | Your key message is missing or wrong | Partly there | Headline or opening paragraph carries your angle | | Spokesperson | No one quoted | Generic company quote from the release | Founder quoted from an actual interview | | Durability | Behind a fast-scrolling feed, gone in a day | Normal article page | Linked, indexed and likely to be cited later by search and AI assistants | Then calculate two numbers for the report as a whole: 1. **Earned share:** earned lines divided by total lines. This tells you how much of the report is actually coverage. 2. **Average earned score:** sum of earned scores divided by number of earned lines. This tells you how good that coverage is. A report with 150 lines, 6 of them earned with an average score of 7, is a very different month from what the cover slide claims. A report with 9 lines, all earned, averaging 6, might be the better month. That's the point of the exercise. ## A worked example from my own reporting I apply the same buckets to my own clients' reports, and I'd rather show where my own work lands than pretend everything is earned. For Bullieverse's $4M raise in February 2022, the scope included the funding release, organic pitching and some sponsored placements. Two lines in that report are explicitly not earned. The CoinDesk placement was sponsored content, so it goes in the sponsored column. The Cointelegraph line was a press release, so it's syndicated. The organic pitching shows up in Indian business and startup media, Economic Times, Inc42, Entrackr, YourStory, VCCircle and Tech in Asia, plus a Hackernoon founder interview. You can see the full campaign in the [Bullieverse case study](/work/bullieverse). Sorted that way, the report is still strong, and it's honest. A founder reading it knows exactly which outlets chose the story and which ones were paid. The sponsored lines weren't a mistake; they were a scoped decision for a raise that needed reach in a short window. They just aren't coverage. ## Questions to send the agency after the audit Once you've sorted the report, you'll have questions. Send these requests in writing, politely, and expect clear answers. - Mark every line that was paid for, and say whether the cost sat inside the retainer or was billed separately. - For each earned line, name the reporter pitched, the angle used and the date. - Share how many reporters were pitched in total this month, and how many said no. - Name the wire or distribution service behind the syndicated lines. - Outline the plan to raise the earned share next month. The third request matters more than founders think. Earned coverage comes from relationships, and a report that shows only outcomes hides whether any relationships are being built. > lot of PR folks are only focused on driving coverage, only a handful focus on building genuine media relationships > > [Shilika Jain on X, 25 Jun 2026](https://x.com/Shilika_jain/status/2070043456257532411) An agency that's doing the relationship work will happily show you the pitch log. One that isn't will talk about "impressions". ## What a good report looks like after the audit A good report doesn't have to be long. It separates the buckets on the page so you never have to do this sorting yourself. That's how I structure reporting on my own [Web3 PR campaigns](/services/web3-pr-campaigns). It lists earned stories first with the reporter, outlet and angle. It puts paid placements in their own section with the cost. It treats syndicated pickups as distribution, a single line with a count. And it tells you what happened in the pitching, including the stories that didn't land. Read that way, a report with nine earned lines can tell you more than one with ninety mixed ones. If you want the metrics layer on top of this, [how to measure PR results](/blog/how-to-measure-pr-results-startups) covers share of voice, referral traffic and AI citation tracking. The audit here is the step before that: making sure you're measuring the right lines. ## When this audit is the wrong tool Two situations where I'd skip or soften it. **You bought distribution on purpose.** If your contract was for paid placements or wire distribution, and that's what you wanted, the audit will just confirm it. Judge that spend on reach and cost, not on earned share. **Your agency is two months in.** Early reports are thin for honest reasons. Reporters need time to know a new client. Audit the pitch log and the angles instead of the outcomes, and run the full four-bucket sort from month three. There's also a cost to running it. It takes a few hours the first time, and it can sour a relationship if you treat it as an accusation. Frame it as "help me understand the report", not "prove this". If the agency's answers are clear, you've lost nothing. If they aren't, you've learned what you needed to know. For the wider list of warning signs, see [PR agency red flags](/blog/pr-agency-red-flags). ## FAQ ### How long does a coverage report audit take? For a monthly report of 50 to 200 lines, plan on two to four hours the first time, mostly opening URLs and checking bylines and labels. Once you know which domains are syndication partners, later months go much faster. ### Should sponsored coverage ever count toward PR results? Count it as paid media, with its cost next to it. It can be a reasonable spend for reach in a specific market or around a specific moment. It just shouldn't be added to the earned total or reported as coverage a journalist chose. ### What earned share should a crypto PR report have? There's no universal number, and I'd be suspicious of anyone who quotes one. What matters is that earned lines are clearly separated, that the count matches the angles pitched, and that the trend over three months is going up. ### Can I ask an agency to stop including wire pickups in reports? Yes. Ask for them as a single summary line with a count and the distribution service used, below the earned section. Most good agencies will be relieved to report that way. *Holding an agency report you're unsure about? [Send it over](/contact) and I'll sort it into the four buckets with you.* --- **Book a 30-min coverage report audit with Shilika** - https://calendly.com/shilikajain/30min/ Canonical: https://www.shilikajain.com/blog/pr-report-audit-paid-links-vs-earned-coverage