The fastest way to lose a PR budget is to defend it with a number nobody believes. You land a write-up, the report boasts an "advertising value" in the tens of thousands and millions of "potential impressions," and the person holding the budget quietly files it under marketing theater. Meanwhile the genuinely useful signals — who linked, who searched your name afterward, which deals mentioned the article — go unmeasured because they're harder to put on a slide.
The takeaway up front: how to measure PR honestly comes down to one rule — judge it by the outcomes it's supposed to move, not by the size of the audience it theoretically touched. Reach and advertising-value numbers are easy to produce and impossible to defend — they count exposure that may never have happened and price it at a rate PR never paid. Real earned media measurement throws those out, ties each campaign to one business outcome, and reports the change against a baseline set before you pitched.
Why the standard PR metrics lie
Two numbers dominate PR reporting, and both mislead. Advertising Value Equivalent (AVE) estimates what your coverage would have cost as paid advertising, then reports that figure as value delivered. It's been the default for decades and discredited for almost as long — the standards bodies for PR measurement explicitly call for retiring it. The logic breaks twice: an editorial mention and a paid ad aren't equivalent units (one is a journalist's judgment, the other is space you bought), and you never spent that money, so quoting it as "value" invents a cost to claim a saving. Worse, AVE rewards volume over relevance — a throwaway mention in a huge outlet "values" higher than a deep feature in the trade publication your buyers actually read.
Potential reach (or "potential impressions") is the other offender: the summed audience of every outlet that ran your story. The word doing the lying is potential — it counts a publication's entire circulation as if every reader saw your piece, when realistically a sliver did. It's the distribution-side "reach" myth carried into the report; the same trap shows up when buying distribution, which the press release distribution guide walks through. A figure you can't verify and can't tie to an outcome isn't a metric — it's decoration.
The test for any PR metric: could I defend this to a skeptical CFO who asks "how do you know?" AVE and potential reach fail it; almost everything below passes.
Start with the outcome, not the coverage
Measurement goes wrong when it starts at the clipping and works backward, hunting for a number to attach. Start at the other end: what is this PR meant to do? PR usually serves one of a few goals, and each implies different evidence:
- Awareness — more of the right people know you exist. Evidence: branded search volume, direct traffic, share of voice.
- Credibility — you look like a real, trusted player. Evidence: coverage in outlets your buyers respect, validation cited in sales conversations, your "as seen in" proof.
- Demand — coverage moves people toward becoming customers. Evidence: referral traffic, the conversions it drives, deals that name coverage as a touchpoint.
- Reputation defense — protect the brand under scrutiny. Evidence: sentiment trend, share of favorable voice, correction of the record.
Pick the one goal a campaign is actually for before it runs — a funding announcement is usually awareness and credibility; a customer-story push is demand. Trying to make one campaign prove all four is how reports turn into noise. One goal, one or two metrics, measured honestly, beats a dashboard of everything.
The metrics worth tracking
Each of these maps to a real outcome and can be sourced, not guessed.
Coverage quality, not count
A tally of "37 placements" says nothing about whether any mattered. Grade coverage instead. For each piece, note relevance (does this outlet reach your buyers?), prominence (a dedicated feature, or a one-line aside in a roundup?), and message pull-through (did it carry your key message, or just your name?). A handful of high-relevance pieces that land your message beats dozens of stray mentions, and reframes the report around value, not volume.
Share of voice
Share of voice is your slice of the conversation: your brand's mentions as a percentage of your category's total. It's useful precisely because it's relative — it controls for a slow news month and shows movement against competitors, which a raw count never does. Track the trend, not a single reading, and pair it with sentiment so you're measuring favorable share, not just loud share.
Referral traffic and what it does next
This is where earned media meets analytics. Coverage with a link sends real, attributable visitors; tag them and follow them past the session to an action: signups, demo requests, subscriptions, purchases. A small stream that converts beats a big one that bounces — though plenty of valuable coverage carries no link, so treat referral traffic as one input, not the whole verdict.
Branded search and sentiment
When PR works, more people go looking for you by name. A lift in branded search queries and direct traffic right after a campaign reflects intent — someone heard about you and acted — which makes it one of the cleaner awareness signals available; watch the bump against your baseline and line it up with publish dates. For reputation work, add sentiment (favorable, neutral, negative), tracked by reading the coverage against a consistent rubric.
Build a baseline, then report against it
Every metric above is only meaningful as a change. A branded-search count is trivia; "branded search up versus the prior three-month average, right after the campaign" is evidence. So before the next push, write down where things stand — branded search, referral traffic, share of voice, sentiment — then measure the delta against publish timing. Stay honest about attribution, too: PR rarely acts alone, so don't credit it with every good outcome, and don't demand ad-style last-click proof from a channel that works upstream of the click.
You don't need a measurement suite to report this well — one page does it: the goal, what ran (coverage graded, not counted), the two or three numbers that map to the goal against their baseline, and what it plausibly drove. No AVE, no potential reach, nothing you couldn't defend to the person signing the checks. A report that admits its limits is the one that earns trust — and the budget — over time.
FAQ
Why shouldn't I use Advertising Value Equivalent (AVE)?
Because it measures the wrong thing and inflates it. AVE prices your coverage as if it were paid advertising you bought, but editorial coverage and paid ads aren't equivalent, and you never spent that money — so the "value" is invented. It also rewards big-outlet volume over the relevant coverage real buyers read. The PR measurement standards bodies have called for retiring AVE for exactly these reasons.
How do I show PR ROI without sales attribution?
Set a baseline before the campaign and report the change after it across signals that point at value: branded-search lift, referral traffic and what it does on your site, share-of-voice movement, and graded coverage quality. You won't get clean last-click ROI from PR, and shouldn't pretend to — it works upstream of the click. A documented baseline, a visible delta, and honest language about contribution beats a fabricated dollar figure.
How long before PR results show up in the metrics?
Awareness signals like branded search and referral traffic often move within days of coverage going live, so watch the window right after each publish date. Slower-building outcomes — share-of-voice trends, sentiment shifts, and any downstream pipeline effect — play out over weeks and months. Measure both the immediate bump and the longer trend, and judge a PR program on the cumulative line, not a single placement.
Next step
Pick one campaign you can see coming and one outcome it's genuinely for. Write down the baseline now — branded search, referral traffic, share of voice, sentiment — so you have something to measure against. After it runs, report the delta in plain language, grade the coverage instead of counting it, and leave AVE and potential reach out entirely. The goal isn't a prettier number; it's one you can defend when someone asks how you know PR is working. For more on earning the coverage worth measuring, start at prwhero.com.
