I have run PR programmes at half a dozen agencies and inside a large one, and I will say this plainly: most PR measurement is theatre. The clearest evidence comes every time we take over an account from an outgoing agency. The reporting is wildly inflated, almost always in the same way — press releases and their automatic syndication counted as equivalent to organic earned coverage, rolled into one big number designed to look like momentum. Then the new agency arrives, counts honestly, and the number collapses. Nothing changed except who was doing the arithmetic.
How the inflation works
It is rarely outright dishonesty. It is a series of choices that all point the same direction.
Syndication counted as coverage. One press release republished across 200 aggregator sites becomes "200 placements". Nobody read 199 of them. This is the single largest source of inflated numbers in the industry.
Press releases counted as earned media. You wrote it, you paid to distribute it, and it ran verbatim. That is owned content on someone else's domain. Counting it alongside a journalist choosing to write about you conflates two completely different things.
Impressions summed across outlets. Each publication's total audience, added together, produces figures larger than the population of most countries. The number is not wrong so much as meaningless.
Advertising value equivalency. Assigning coverage the notional cost of buying equivalent ads. The industry's own professional bodies have advised against it for years. It survives because it converts a modest month into a large dollar figure.
One aggregate number. Reach, mentions and AVE combined into a single headline metric. Aggregation is what makes the inflation invisible — you cannot see which component is carrying it.
The underlying incentive: it is easier to prove you were busy than to prove you were effective. Most agency reporting optimises for demonstrating activity, because activity is what the agency controls.
What we do instead
Not a claim about intent — a description of what the client actually receives.
A weekly coverage report, automated, delivered into a shared channel the client sits in. Not a monthly deck assembled by an account executive the night before a call. It generates every week, it goes to the client directly, and nobody hand-picks what goes in it.
Competitor tracking alongside it. Your coverage means little without knowing what the category looked like that week. If everyone got quiet, your quiet week is not a failure. If a competitor had three placements you did not, that is worth knowing on Monday rather than at the quarterly review.
Built from scratch, because the tools do not do it. We ran on Muck Rack and are not renewing. It is a competent product but it could not answer the questions we needed answered, so we built our own tracking and are replacing it at roughly a quarter of the subscription cost.
Firehose for what traditional monitoring misses. Real-time crawling that catches mentions in places media monitoring tools do not index well.
Social measured directly through the API. We pull X mentions ourselves rather than accepting a platform's summary, which lets us see genuine week-over-week spikes and tie them to specific coverage.
No inflated numbers. Syndication is labelled as syndication. Press releases are labelled as press releases. Organic earned coverage is counted separately, because it is a different thing.
The honest part: attribution is genuinely hard
Everything above is about not lying. Tying PR to revenue is a different problem, and I do not think anyone has solved it — including us.
The priming problem. A prospect reads an article about you in March. In May they see a retargeting ad and click it. Analytics credits the ad. The ad worked because the article existed, and nothing in any attribution model captures that.
The due diligence problem. Somebody is about to recommend a large contract with you. Their job is partly on the line. They search your company, and so does their boss. What comes back — coverage in publications they respect, or nothing — affects a decision worth far more than any campaign. It is invisible to every measurement system in use.
The absence problem. You cannot measure the deal that did not stall, or the customer who did not churn because the category coverage made you look permanent.
The new one: AI summaries. Increasingly a buyer asks an assistant about your category before they ever reach a website. The answer is assembled from published coverage. Whether you are named in it is now a material commercial outcome, and no PR reporting tool measures it at all.
Any agency claiming a clean revenue attribution model for earned media is overselling. The honest position is that PR does things that are real and hard to count, and the response to that should be careful measurement of what can be counted — not an inflated number standing in for the rest.
What we actually track
Reply rate, in your category. Expect 8–11% on a well-built list.
Positive reply share. 63% a yes, 19% not interested, across 175 graded replies. Low reply rate with high positive share is a distribution problem. Low positive share is a story problem. These need different fixes.
Repeat contact rate. How often reporters come back unprompted. The strongest signal a programme is working and the hardest to fake.
Time to quote approval. Under four hours keeps you in the stories that matter. Two days excludes you structurally. Not in any textbook and it decides more than most metrics.
Coverage in the outlets your buyers read, listed individually with the reason each one matters. Never as a volume figure.
Whether AI answers name you. Ask the major assistants monthly and record it. Twenty minutes, and no tool does it yet.
The failures. Twelve of our 522 reply messages were classified "confused" — the reporter could not tell what was being pitched. Smallest number in the dataset and the most actionable, because it points at a defect rather than a matter of taste.
"We log every reply and classify it. Positive, neutral, decline, confused. 'Confused' is the most useful one, because it means the pitch failed before the reporter even got to whether they cared."
Kristin Kaplan, Operations, Salient PR
Questions to ask your current agency
"How many of these placements are syndication?" If the answer is not immediate, the number is inflated.
"Which of these did a journalist choose to write, and which did we write ourselves?"
"What is our reply rate, in our category?" If nobody has it, nothing upstream of coverage is being measured.
"Show me the pitches that failed."
"Ask for the last 20 pitches they sent in your category and what came back. Not case studies, not a logo wall. The actual sends. If they can't produce that, they aren't measuring, and you'll find out in month four."
Justin Mauldin, Founder, Salient PR
"Can I see the raw data, not the deck?" The deck is the interpretation. The data is the thing.
A note on open rate
Every outreach tool reports it and the number is unreliable. Tracking pixels only fire if the recipient's client loads remote images, and privacy-first clients block or proxy them by default.
"Our recorded open rate is 51%. The real number is closer to 7 in 10, because a big share of reporters read mail in clients that block tracking pixels. We know it happens because we get replies from people our system says never opened the email. Treat any open rate you're quoted as a floor, including ours."
Justin Mauldin, Founder, Salient PR
What a usable report contains
One page:
- Pitches sent, replies, reply rate — against last month
- Positive share of replies
- Coverage placed, by outlet, with syndication listed separately and labelled
- Reporters who came back unprompted
- Time to quote approval, and any deadline missed because of it
- Where you appeared in AI answers about the category
If it is longer than a page, it is probably measuring activity.
The short version
- Most PR reporting is inflated in the same way: syndication and press releases counted as earned coverage, rolled into one big number.
- Taking over an account is where you see it — the number collapses under honest counting.
- Attribution is genuinely hard. Priming, due-diligence searches and AI summaries all matter and none are measurable.
- Anyone selling you clean revenue attribution for earned media is overselling.
- Track reply rate, positive share, repeat contact, approval speed, outlets that matter, and AI mentions. Ours: 63% of graded replies a yes.
- Ask what proportion of your placements is syndication. The pause before the answer tells you most of it.
Salient PR delivers automated weekly coverage reports with competitor tracking directly into a shared channel with each client, built on our own tooling rather than an off-the-shelf suite. All figures come from our outreach database covering December 2025 to August 2026 and are updated quarterly.
Sources
- Salient PR outreach database, Dec 2025 – Aug 2026
If you want reporting that shows the denominator rather than a headline number — Salient PR is a technology PR agency working with technology companies. See how we work or start a conversation.
We also publish what a yes actually turns into — nine pitches traced from the reporter's reply to the published piece, with the dates.
Related reading
About Salient PR
Salient is a B2B tech PR agency. We help startups become the story — landing tier-one coverage, building executive visibility, and turning attention into pipeline.
Senior team. No hand-offs. Radical transparency. If you’re ready to talk, reach out.
Which PR metrics mean something, which are theatre, and what to ask for when you take over an account.
