How to measure press placement ROI
The metrics that matter • the vanity ones that do not • a simple model
PR has a measurement problem: the industry reports impressions and “ad value equivalency”, which are close to meaningless. Here is what to actually track, how to attribute a placement honestly, and a model you can run in a spreadsheet.
Why PR reporting is mostly theatre
Two numbers you should refuse to accept.
Impressions are usually the publication’s total monthly audience, presented as though every one of those people read your article. They did not. A site with 100 million monthly visitors did not deliver 100 million impressions to your piece.
Ad Value Equivalency (AVE) is worse. It multiplies your article’s column inches by the publication’s advertising rate to produce a “value”. The industry’s own professional bodies have spent years telling practitioners to stop using it, because editorial coverage and advertising are not interchangeable, and the number can be inflated arbitrarily.
If an agency reports impressions and AVE, they are not measuring anything. They are producing a large number to justify a retainer. Ask instead what changed in your business.
Six metrics that actually mean something
All of them are measurable with tools you already have.
| Metric | How to measure it | What good looks like |
|---|---|---|
| Branded search volume | Search Console: impressions and clicks for queries containing your brand or founder name, 30 days before vs 30 days after | A sustained lift, not a one-day spike |
| Direct traffic | Analytics: direct and unattributed sessions in the two weeks after publication | A visible step change, then a higher baseline |
| Close rate | CRM: percentage of qualified opportunities won, before vs after | Even 2–3 points is large money on a real pipeline |
| Sales-cycle length | CRM: median days from first contact to closed-won | Credibility usually shortens the cycle before it lifts the rate |
| AI citations | Ask ChatGPT, Perplexity and Google AI Overviews about your company monthly and record what they say and cite | Being described accurately, and cited to a real source |
| Referral traffic | Analytics: sessions from the publication’s domain | Small but high-intent. Judge on conversion rate, not volume |
A four-step model you can run in a spreadsheet
Honest attribution without pretending to certainty.
Baseline before you buy
Record 90 days of branded search, direct traffic, close rate and median cycle length. Without a baseline you cannot measure anything, and most people skip this step and then argue about results.
Mark the publication date
Annotate it in your analytics. Compare the 30 and 90 days after against your baseline, and note anything else that changed in the same window — a campaign, a launch, a seasonal swing.
Value the movement conservatively
Take the lift in close rate, multiply by opportunities and average deal size. Then halve it, because you cannot prove the placement caused all of it. A defensible half beats an indefensible whole.
Add the asset value
The article is reusable forever — in decks, ads, bios, petitions and your site. Value it as what you would have paid to produce equivalent third-party credibility, then judge the total against the placement cost.
What placement ROI cannot tell you
Four limits worth stating out loud.
Attribution is never clean
Buyers see a logo on your site, read the article months later, and mention it in negotiation. No analytics tool captures that path, and anyone claiming precise attribution is inventing it.
The value is back-loaded
Entity and search effects build over months. Judging a placement at 14 days is like judging a hire in their first week.
One placement is a small sample
A single article rarely produces a clean measurable step change. A programme of coverage does. Budget accordingly, and measure over quarters.
Some value is defensive
A credible article that prevents a deal from stalling never appears as a gain. It shows up as an absence of loss, which no dashboard reports.
Placement ROI FAQ
What is a realistic ROI on a press placement?
Why do agencies report impressions?
Is AVE ever legitimate?
How long should I wait before judging a placement?
How do I measure AI visibility?
What if I see no measurable lift?
Do you report on this for clients?
Buy coverage you can measure.
Tell us the metric you need to move. We will tell you which outlet moves it, and how to know whether it worked.