Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

PR & Earned Media › PR Measurement

PR KPIs: Outputs, Outcomes and Impact

How to choose PR KPIs: separating outputs, outcomes and impact, coverage quality over volume, why advertising value equivalence misleads, and setting targets.
Guideintermediate8 min readUpdated

Why vanity coverage counts mislead

The easiest thing to report in PR is a count: pieces of coverage, total audience of the titles, mentions. Counts are attractive because they always go up and are simple to explain. They mislead in several ways.

A count treats every piece as equal. A detailed feature in the one trade title that buyers read and a passing mention in a syndicated round-up both score one. It ignores whether the coverage carried the intended message, or any message, and whether it was positive. Audience totals compound the problem, because the readership of a site is not the number of people who saw a particular article, and adding them across titles double-counts readers who see several. And counts say nothing about effect: a programme can grow coverage every quarter while awareness, consideration and enquiries stay flat.

The result is a reporting habit that rewards the wrong behaviour. Teams chase volume by pitching widely to irrelevant titles, issuing releases for minor news, and counting syndicated copies. The fix is not to stop counting but to put counts in their place: as outputs, qualified by quality, and subordinate to outcome measures. The general principle of tying every KPI to an objective, set out in the objectives and KPIs guide, applies to PR as much as to any channel.

Outputs, outcomes and impact

The ladder gives each kind of measure its proper role. The PR measurement hub introduces it; this section shows how to use it to choose KPIs.

LevelWhat it measuresGood measuresWeak measuresRole
OutputsThe coverage and its qualityPieces in target titles, message inclusion, prominence, spokesperson quoted, links from relevant titlesTotal pieces, total title audience, syndicated copiesLeading indicators and diagnostics
OutcomesChanges in the audienceEarned share of voice, awareness and consideration among the target audience, branded search demand, referral visits, sentiment in coverageSocial impressions with no audience filterThe main KPI level for most programmes
ImpactContribution to the businessEnquiries or sales influenced, reputation scores, recruitment response, price or retention effectsRevenue claimed by last-click attributionStrategic justification; reviewed less often

Choose one or two KPIs at the outcome level, supported by a small set of output measures that explain them, and at least one impact measure or line of evidence reviewed periodically. More than that becomes a dashboard nobody reads. For the person who ran a programme and must write up their own part in it, the guide to showing brand, PR and community results shows how to keep outputs, outcomes and impact in separate wording.

Quality of coverage: relevance, message, prominence

Coverage quality can be scored consistently with a few fields, so that outputs become informative rather than merely countable. Three dimensions carry most of the weight.

Relevance is whether the title and the piece reach the audience the objective names. A tier list of target titles, agreed in advance, makes this objective: coverage in a top-tier title counts more than coverage elsewhere. Message is whether the piece carried one of the agreed key messages, quoted a spokesperson, or described the organisation accurately. A piece that mentions the brand but not the point has limited value. Prominence is how visible the organisation is in the piece: headline, opening, a passing mention, or an image. Some teams add sentiment and whether the piece included a link.

A simple scoring scheme across these dimensions, applied by the same people in the same way, turns a coverage log into a quality-weighted output measure. It also makes the outcome measure of earned Share of Voice more meaningful, because a share calculated on quality-weighted coverage reflects what readers actually encountered. Scoring is a judgement, so write down the rules and check a sample periodically to keep it consistent. The routine for logging these fields is covered in the media monitoring and coverage reporting guide.

Advertising value equivalence and why to drop it

Advertising value equivalence, usually shortened to AVE, prices coverage as if it were paid space: the area or duration of the piece is multiplied by the advertising rate for that space, sometimes with an arbitrary multiplier for the supposed extra credibility of editorial. It produces a large, precise-looking number that has been used for decades to justify PR budgets.

The problems are fundamental. Coverage is not advertising: the organisation did not control the message, the placement or the tone, so the price of advertising space is not its value. AVE ignores whether coverage was relevant, positive or seen by the right people; a critical article scores as highly as a favourable one. The multipliers have no evidential basis. And the number invites a misleading comparison with media budgets, implying that PR "returns" a multiple of its cost when it has measured nothing of the sort. Industry measurement bodies have argued against AVE for many years, and effectiveness evidence from bodies such as the IPA focuses on outcomes and business effects rather than on notional media value.

Dropping AVE is easier with a replacement ready. Offer the board outcome measures it can understand, such as share of voice in target titles, awareness among buyers and branded search trends, and show how they connect to business objectives through the KPI tree described next.

A PR KPI tree

A KPI tree links a business objective at the top to the outcomes that drive it and the outputs that should move those outcomes at the bottom. It forces explicit choices about what PR is expected to change. The general method is covered in the KPI trees guide and the KPI tree framework; the widget below lets you build one.

For PR, the outcome layer usually includes awareness or consideration among the target audience, earned share of voice and branded search. The evidence from the Ehrenberg-Bass Institute on how brands grow, mainly by reaching more buyers and being easy to think of in buying situations, is a useful reason to put broad Reach among relevant buyers and mental availability near the top of the outcome layer rather than narrow engagement measures. The framework of mental and physical availability explains how that outcome can be measured with situation-cued recall among category buyers. Measuring awareness and lift properly is covered in the brand awareness and lift guide.

Illustrative example — not a real result

A KPI tree for a business-to-business PR programme

Cobblewren Fleet Solutions, which leases vans to small trade businesses, sets a business objective of growing new contracts by 15% in a year. The tree beneath it has three outcome KPIs for PR. First, prompted awareness among small trade business owners, measured twice a year by a survey of 500, with a target of moving from 22% to 28%. Second, earned share of voice across 12 trade and small-business titles, with a target of moving from 10% to 16%. Third, monthly branded search volume. Beneath these sit output measures: 24 pieces of coverage in tier-one titles over the year, key message carried in at least 60% of them, and the fleet director quoted in 10. Leads mentioning an article are logged by sales as supporting evidence for impact.

Read the tree from the bottom when diagnosing and from the top when planning. If outputs are on track but an outcome is flat, the coverage is probably in the wrong titles or missing the message, and the fix sits in the media list or the story. If outcomes move but the business objective does not, the problem may lie outside PR, in the product, the price or the sales process, and the tree makes that conversation easier to have honestly. Keep the tree small: a single objective, a few outcomes, and only the outputs that plausibly move them. Share it with the analytics and brand teams so the same outcome measures appear in their reports, which stops PR being judged on a separate set of numbers that nobody else recognises.

Build a PR KPI tree

Targets and review cadence

Targets need a baseline. Before setting one, measure the current position for each KPI over a meaningful period: earned share of voice over several months, awareness from at least one survey wave, branded search over a full year to see seasonality. Without a baseline, any target is a guess, and any result can be presented as success or failure. Before a new programme, the tutorial on running a PR audit codes past coverage and compares it with competitors to produce that starting point.

Set targets on outcomes as ranges rather than single figures, because outcome measures are noisy and PR's effect builds over time. Set output targets that are plausibly sufficient to move the outcomes, and treat them as plans rather than goals in their own right. Agree in advance what would count as evidence of impact, such as a lift in enquiries following campaigns, or a difference between regions with and without activity, and accept that click-based Attribution will undercount PR.

Cadence should follow how fast each measure moves. Outputs can be tracked continuously and reviewed monthly to catch problems early. Outcomes such as share of voice and branded search are best reviewed monthly or quarterly against a trend. Awareness and impact measures move slowly and are reviewed quarterly or twice a year. Revisit the targets themselves once a year, or when the business objective changes. The PR programme playbook builds these reviews into its monthly coverage report and quarterly story bank review.

When the objective is awareness, PR for brand awareness applies these outcome measures to a continuous programme of data stories and expert comment judged on share of voice.

When the objective is consideration, PR for buyers weighing their options tracks inclusion and factual accuracy as output measures across a fixed set of the reviews, comparisons and roundups those buyers read, and reads them beside referral visits and what new customers say helped them choose.

Common mistakes

  • Reporting coverage counts as the headline. Counts are outputs; put outcome measures first.
  • Adding up title audiences. Readership of a site is not the audience for a piece, and totals double-count.
  • Using AVE. It prices coverage as advertising and measures nothing about effect.
  • No quality scoring. Without relevance, message and prominence fields, outputs are uninformative.
  • Targets without a baseline. Measure the starting point before setting where to go.
  • Too many KPIs. One or two outcome KPIs with a few supporting outputs are easier to act on than a long dashboard.
  • Expecting attribution to show PR. Click-based models will undercount earned media; plan other evidence.
  • Reviewing slow measures weekly. Awareness and reputation move slowly; frequent review mostly measures noise.

Frequently asked questions

What is the single best KPI for PR?

There is no single best one, because the right KPI depends on the objective. For most programmes aimed at awareness, earned share of voice in target titles combined with a measure of audience awareness or branded search is a sound core. Coverage counts should be tracked as an output, not used as the headline KPI.

Why should we stop using advertising value equivalence?

It prices coverage as if it were advertising space, which it is not, and it ignores whether the coverage was relevant, positive, seen by the right people or did anything. Industry measurement bodies have long argued against it. A board that hears a large AVE figure learns nothing about what PR achieved and may compare it wrongly with media budgets.

How often should PR KPIs be reviewed?

Track outputs continuously, review outcomes monthly or quarterly depending on how quickly the data moves, and review impact and targets once or twice a year. Outcomes such as awareness change slowly, so judging them every week mostly measures noise.

Terms used on this page

Frameworks

Where this page is used

Also related

Sources

  1. Institute of Practitioners in Advertising — IPA
  2. Ehrenberg-Bass Institute for Marketing Science — Ehrenberg-Bass Institute

Related