Why PR measurement is hard
Paid media can be measured from the inside: the platform reports impressions, clicks and conversions it delivered. Earned media cannot. Coverage appears on sites the organisation does not run, is read by people it cannot identify, and is often seen without a click. The effect that matters, such as a buyer trusting the brand more or remembering it when a need arises, happens in people's heads and shows up later, through other channels.
Three consequences follow. First, PR's outputs are easy to count and tempting to report, so programmes drift towards coverage volume, which says little about effect. Second, the behaviours PR influences, such as searching for the brand, visiting the site directly or accepting a sales call, are usually credited to other channels by any click-based AttributionAssigning conversion credit to touchpoints. model. Third, PR's effects build over time, so judging it campaign by campaign makes it look erratic. Published effectiveness evidence, such as the IPA's work on how communication builds brands over longer periods, supports judging earned media on trends in outcomes rather than on short-term tallies.
None of this makes PR unmeasurable. It means measurement has to combine coverage data, audience research, search and site data and, where possible, business results, and has to be honest about which links in the chain are evidence and which are inference. The PR and earned media hub places measurement at the end of the PR system, where it also feeds the next round of planning.
Outputs, outcomes, impact
The most useful frame for PR measurement is a three-level ladder. Each level answers a different question, and a programme needs at least one measure at each.
| Level | Question | Example measures | Source |
|---|---|---|---|
| Outputs | What appeared, where and saying what? | Coverage in target titles, message inclusion, prominence, spokesperson quoted, links earned | Monitoring and the coverage log |
| Outcomes | What changed in the audience? | Awareness, consideration, earned share of voice, branded search demand, referral visits | Surveys, search data, analytics, monitoring |
| Impact | What did it contribute to the business? | Enquiries and sales influenced, reputation scores, recruitment, cost of acquisition | Business data, controlled tests, modelling |
Outputs are necessary but not sufficient: without coverage nothing happens, but coverage alone proves nothing. Outcomes are where PR should usually be held to account. Impact is the hardest to show and the most persuasive when it can be, typically through trends that line up with campaign timing, comparisons between regions or periods, or the organisation's wider modelling. A PR KPIA metric tied directly to an objective. should sit at the outcome level wherever possible, with outputs tracked as leading indicators.
The core guides
PR KPIs: outputs, outcomes and impact covers how to choose KPIs: why coverage counts mislead, how to judge coverage quality by relevance, message and prominence, why advertising value equivalenceA measure that prices press coverage as if it were paid space, multiplying its size or duration by the advertising rate; the PR guides say to retire it. should be dropped, how to build a PR KPI tree from a business objective, and how to set targets and a review cadence.
Media monitoring and coverage reporting covers the routine that supplies the data: what to monitor and where, the fields a coverage log needs, how to calculate earned share of voice from coverage, how to read referral, link and search effects, and how to build a report that people outside PR will read.
Share of voice as a general brand metric, including its paid and search forms and its relationship to market share, is owned by the share of voice guide in the branding section; the guides here use its earned form and link back for the definition.
Further guides
The further guides cover the groundwork that comes before the first target is set:
- Running a PR audit before a new programme — a tutorial for reading past coverage for titles, themes and tone, competitors' share, the media list, materials and what journalists think, written up as a baseline.
PR in the marketing measurement plan
PR measurement works best as part of the organisation's wider measurement plan rather than as a separate report. That plan sets out the business objectives, the marketing objectives beneath them, the KPIs for each, and the data sources and owners. PR's place in it is usually at the awareness and consideration levels, alongside brand advertising and social, with shared measures such as ReachUnique people who saw an ad. among the target audience, brand awareness and branded search.
A KPI tree makes the connection explicit: a business objective at the top, the outcomes that drive it in the middle, and the PR outputs that should move those outcomes at the bottom. Drawing the tree with colleagues from analytics and brand forces agreement on which outcomes PR is expected to move and how they will be observed, before the first campaign rather than after it.
Attribution needs particular care. Click-based models will undercount PR because most of its effect does not arrive through a tracked click. The analytics team owns attribution mechanics; the PR team's job is to make sure the plan includes measures that can see PR's contribution, such as branded search trends, direct traffic, survey questions on where people heard of the brand, and, where volume allows, comparisons between periods or regions with and without activity.
Putting PR into the measurement plan
Oakridge Software, which sells scheduling tools to clinics, adds PR to its annual measurement plan for the first time. The business objective is 20% more qualified demos. The plan assigns PR two outcome KPIs: earned share of voice in 8 healthcare trade titles, and branded search volume, with coverage in target titles as a leading output. A quarterly survey of 300 clinic managers adds a question on brand awareness. After three quarters, earned share of voice rises from 12% to 21%, branded searches rise in the months following each campaign, and the awareness question moves from 18% to 26%. The analytics team cannot isolate PR's share of the demo growth, but the trend gives the board enough to fund the programme for another year.
Common PR measurement problems
| Problem | Usual cause | Guide |
|---|---|---|
| Reports list coverage but nobody acts on them | Measuring outputs only; no outcome KPIs | PR KPIs |
| Leadership asks what the coverage was worth | No agreed outcome measures; temptation to use advertising value equivalence | PR KPIs |
| Targets feel arbitrary | No baseline; targets set on volume rather than outcomes | PR KPIs, Running a PR audit |
| Coverage is missed or counted twice | Patchy monitoring; no consistent log | Media monitoring and coverage reporting |
| Share of voice moves unpredictably | Inconsistent competitor set, sources or period | Media monitoring and coverage reporting |
| Web effects of coverage are invisible | No referral tracking or link log; branded search not watched | Media monitoring and coverage reporting |
| PR gets no credit in attribution reports | Click-based models undercount earned media | PR KPIs |
