PR measurement has a credibility problem, and it is self-inflicted. For decades the industry reported "12 million impressions" and "$840,000 in ad value equivalency," numbers that sound impressive and mean almost nothing. Every finance leader who has ever asked "so what did that get us?" already knows those figures are theater. The hard problem in PR measurement is not producing big numbers — it is producing honest ones that connect coverage to outcomes a skeptic will accept.
The takeaway up front: stop measuring how far your coverage traveled and start measuring whether it said what you wanted, to the people you wanted, with a result you can trace. That means three things working together — message pull-through (did the coverage carry your point), share of voice measured fairly (are you winning the conversation that matters), and a coverage-to-outcome trail (did the coverage move a real metric). Get those three right and you can defend your program in any budget meeting. This guide shows how.
Why the standard metrics fail
Two numbers dominate bad PR reports, and both are broken for the same underlying reason: they measure potential and call it impact.
Reach / impressions counts how many people could theoretically have seen the coverage — usually the outlet's total monthly traffic, not the readers of your specific article. A mention buried in paragraph eleven of a piece on a site with 10 million monthly visitors does not give you 10 million impressions. It gives you the few thousand who read that piece and the few hundred who reached your paragraph. Reporting the 10 million is not optimism; it is fiction.
Advertising Value Equivalency (AVE) is worse. It takes the column inches or seconds of your coverage and multiplies by what an ad of that size would cost, as if a journalist's independent mention were equivalent to a paid ad you wrote yourself. The entire PR industry's standards body, AMEC, has explicitly rejected AVE for years, for an obvious reason: earned media's value comes from third-party credibility, which an ad does not have, and from message control, which an ad has and coverage does not. AVE conflates two things that are not the same and assigns a dollar figure that no buyer would ever actually pay.
The deeper failure: both metrics reward volume, so they quietly push teams toward mass-blasting and wire spam — the exact behaviors that destroy the relationships good PR depends on. When your scoreboard rewards noise, you optimize for noise. Fix the scoreboard first.
The three measurements that actually hold up
Replace the vanity numbers with three measurements that a skeptic can interrogate and that still survive.
1. Message pull-through
Message pull-through asks the only question that separates PR from publicity: did the coverage carry your message, or just your name? A thousand mentions that miss your point are worth less than three that nail it.
Before a campaign, write down two or three key messages — the specific things you want the market to believe (for example: "we are the affordable option for small teams," "we are the security-first choice," "our founder is the category's expert voice"). Then score each piece of coverage on whether it actually carried each message. A simple, defensible scale:
- 2 — Carried clearly: the message appears, attributed and intact.
- 1 — Partial: the topic is there but the framing is weak or diluted.
- 0 — Absent: you were mentioned, but your message was not.
- −1 — Off-message: the coverage actively undercut your positioning.
Pull-through rate = (sum of scores) ÷ (number of pieces × 2). This single number tells you whether your pitching and spokesperson prep are working, and it is the metric that most directly maps to the strategy your leadership actually cares about. Two strong-message placements beating ten name-only mentions is exactly the prioritization you want to reward.
2. Share of voice — measured fairly
Share of voice (SOV) is your slice of the total conversation in your category. The trap is measuring it lazily: raw mention counts across "the whole internet," which rewards whoever spams the most press releases. To make SOV honest, constrain it on three axes:
- A defined competitive set — you versus a named list of 3–6 real competitors, not "the industry."
- A defined surface — the specific outlets and beats that reach your buyers, not every blog that scraped a wire.
- Quality-weighted — weight each mention by tier (a tier-1 trade feature counts more than a syndicated republish) and ideally by pull-through, so on-message coverage counts more than a passing name-drop.
Measured this way, SOV answers a real strategic question: are we winning the conversation in the rooms where it counts? A rising quality-weighted SOV against a fixed competitive set is one of the few PR metrics a board immediately understands.
3. The coverage-to-outcome trail
The hardest and most valuable measurement: connecting coverage to something the business cares about. PR will rarely give you clean last-click attribution, and pretending otherwise is how you lose credibility. Instead, build a trail of corroborating signals:
- Referral traffic from the specific coverage URLs (UTM-tagged links where you control them; referrer data where you don't).
- Branded search lift — direct and branded-search traffic in the days after major placements, the classic fingerprint of awareness PR that has no clickable link.
- Direct/dark traffic spikes correlated to broadcast or print hits that can't pass a referrer at all.
- Assisted conversions — pipeline or signups that touched a coverage URL anywhere in their path, not just as the final click.
None of these alone is proof. Together, time-aligned to your placements, they form a defensible narrative: "the week this feature ran, branded search rose 28%, the article drove 1,900 referral sessions, and 14 of them entered the pipeline." That is a sentence a CFO can act on.
A worked example
A 12-person B2B SaaS company runs a one-quarter PR push around a data study. Old-school report: "47 placements, 31 million impressions, $412,000 AVE." Impressive, useless.
Here is the same quarter measured properly. Of 47 placements, 9 are real target-outlet pieces; the rest are wire republications. Across those 9:
- Message pull-through: scores of 2,2,1,2,0,1,2,1,2 = 13 out of a possible 18 → 72% pull-through. Strong: most coverage carried the "affordable for small teams" message intact.
- Share of voice: within a fixed set of 4 competitors across 12 target trade outlets, quality-weighted mentions give the company 34% SOV, up from 19% the prior quarter — it moved from third to first in its category's conversation.
- Outcome trail: the two tier-1 features drove 3,100 referral sessions; branded search rose 41% week-over-week; 22 assisted conversions touched a coverage URL, of which 5 closed at an average $9k ACV → $45k in influenced revenue traced, with a credible halo beyond it.
Notice the report shrank from 47 placements to 9 that mattered, and got dramatically more persuasive. You are no longer defending a fake dollar figure; you are showing a real one and the mechanism behind it.
Common mistakes — and why people make them
- Reporting reach as impact. People do it because big numbers are easy and flattering, and because nobody asks for the methodology until budgets get cut. Pre-empt it: report read-through-adjusted reach or skip reach as a headline entirely.
- Still using AVE. Teams cling to it because it produces a dollar figure leadership seems to like — until a finance leader asks how a journalist's free mention equals a paid ad and the whole report collapses. Lead with influenced pipeline instead; it's a real dollar figure you can defend.
- Counting volume over message. This is the seductive one: more mentions feel like more success, so teams optimize for blasting. But volume without pull-through is just noise, and chasing it erodes the journalist relationships that produce your best coverage. This is exactly why measurement and relationship-building have to move together — the same discipline that builds trust, covered in our media relations guide, is what produces the on-message coverage your scorecard rewards.
- Demanding last-click certainty. Some teams swing the other way and, unable to prove clean attribution, conclude PR is "unmeasurable" and report nothing. The fix is the corroborating-signals trail — directional honesty beats false precision and beats silence.
Edge cases and caveats
- Awareness-stage PR has no clickable link. A print feature or a podcast can't pass a referrer. Lean on branded-search lift and direct-traffic correlation, and pre-register the baseline week so the lift is visible.
- Crisis and reputational work invert the scoreboard. Here, lower SOV and fewer negative-pull-through pieces are the wins. Don't report a crisis quarter on the same template as a launch quarter.
- Long sales cycles outrun a quarterly report. For enterprise pipelines, measure assisted influence and message pull-through now, and treat closed-revenue attribution as a trailing, annual view.
- Small sample sizes are noisy. With under ~8–10 placements, report the individual pieces and their pull-through scores rather than percentages that imply false precision.
FAQ
What should replace AVE in a PR report?
Lead with influenced or assisted pipeline (a real dollar figure traced through a coverage-to-outcome trail), and support it with quality-weighted share of voice and message pull-through. Together these answer "did it carry our message, are we winning the conversation, and did it move a metric" — which is everything AVE pretended to and never did.
How do I measure PR when there's no clickable link, like print or broadcast?
Use correlation, not clicks. Pre-register a baseline week, then watch branded-search volume and direct/"dark" traffic in the days after the placement. A clear lift time-aligned to a broadcast or print hit is the standard, defensible fingerprint of link-less coverage.
Isn't share of voice just a vanity metric too?
Only if you measure it lazily. Raw mention counts across the whole internet are vanity. SOV becomes meaningful when you fix a named competitive set, restrict it to outlets that reach your actual buyers, and quality-weight each mention — then it answers a real strategic question about who's winning the conversation that matters.
How do I prove PR drove revenue without clean attribution?
You build a trail, not a single proof point: UTM-tagged referral traffic, branded-search lift, direct-traffic spikes, and assisted conversions that touched a coverage URL anywhere in the path. No one signal is conclusive, but time-aligned to your placements they form a narrative a finance leader can act on. Directional honesty beats false precision.
How often should I report PR results?
Match the cadence to the buying cycle. Report message pull-through and share of voice monthly or per-campaign because they're fast signals you can act on, and treat hard revenue attribution as a trailing quarterly or annual view — especially for long enterprise sales cycles where the outcome lags the coverage by months.
Put it into practice
PR measurement stops being theater the moment you change the scoreboard. Drop reach-as-impact and AVE. Score every placement for message pull-through, measure share of voice against a fixed competitive set on the surfaces that reach your buyers, and build a corroborating trail from coverage to a real outcome. Report nine placements that mattered instead of forty-seven that didn't, and you'll never have to defend a fake number again.
Want to report PR in outcomes instead of impressions? See how PR Rush can help.
