What to monitor and where
Monitoring has two jobs: catching the organisation's own coverage, and tracking the wider conversation, including competitors and the issues the organisation cares about. Both need a defined scope, because unlimited monitoring produces more noise than anyone can read.
| Source | What it catches | How to monitor |
|---|---|---|
| Online news and magazines | Most earned coverage, including online editions of print titles | Monitoring service or news alerts on brand, spokesperson and product names |
| Coverage in editions that do not appear online, or appear behind a paywall | Monitoring service with print coverage; titles' own archives | |
| Broadcast | Radio, television and podcast mentions | Monitoring service with broadcast transcripts; producers' confirmations |
| Trade and specialist titles | The coverage that often matters most to business buyers | Direct subscription and reading; alerts where available |
| Social platforms | Shares of coverage, journalist comment, wider conversation | Social listening, or platform search for smaller programmes |
| Search and links | New links to the site and changes in branded search | Search console, link tools, analytics |
| Competitors and issues | Rival coverage and the topics in the field | Same tools, with competitor and topic queries |
Write the monitoring queries carefully: the brand name and common misspellings, product names, spokesperson names, and campaign names, with exclusions for unrelated uses of the same words. Review the queries every few months, because both language and competitors change. Social listeningMonitoring public conversation about a brand, its category, competitors and the problems it solves, away from the brand's own accounts. overlaps with social media reporting, which the social media reporting guide covers from the social team's side. The tutorial on setting up social listening builds and tests the social queries and routes what they find to PR, service and the crisis owner.
Logging coverage: fields that matter
A coverage log is the record from which every output and many outcome measures are calculated. A spreadsheet is enough for most programmes; the discipline lies in logging every piece consistently. The fields below support the quality scoring described in the PR KPIs guide.
| Field | What to record |
|---|---|
| Date and title | Publication date, title, section, writer |
| URL or reference | Link for online pieces; page or programme reference otherwise |
| Tier | The agreed priority tier of the title for the objective |
| Campaign or story | Which story or campaign generated it, or reactive or unprompted |
| Prominence | Headline, lead, substantial mention or passing mention |
| Message | Which key message, if any, appeared |
| Spokesperson | Whether a spokesperson was quoted, and who |
| Sentiment | Positive, neutral or negative, judged consistently |
| Link | Whether it links, to which page, and whether the link is followed |
| Competitors mentioned | Which competitors appeared in the same piece |
| Notes | Errors to correct, follow-up opportunities |
Log as you go rather than in a monthly rush, and record pieces that did not go well as faithfully as those that did. A log that only contains good news cannot explain why outcomes moved or did not. When the notes field records an error or a critical piece, the guide to responding to negative or inaccurate coverage covers deciding whether to respond and asking for a correction. Sentiment and prominence are judgements, so write short rules for each and check a sample from time to time.
Share of voice from coverage
Earned share of voice is the organisation's portion of the coverage in a defined set of titles, among a defined set of competitors, over a defined period. The share of voice guide owns the definition of the metric and its paid and search forms; this section covers calculating the earned form from monitoring data.
Three choices determine whether the number is meaningful. The competitor setThe fixed list of brands a measure such as a brand tracker, share of search or share of voice compares against, chosen from what buyers really consider. should be the organisations the target audience actually considers, fixed for the period so that changes reflect coverage rather than a different list. The title set should be the titles that reach the target audience, usually the top tiers of the media list, rather than all coverage everywhere. The unit can be a simple count of pieces mentioning each organisation, or a quality-weighted count using prominence and sentiment, which reflects what readers saw better than a raw count does.
Calculate it monthly and read it as a trend over quarters. Short-term swings are often caused by a single story, positive or negative, for one organisation. A sustained rise in Share of VoiceYour share of market visibility. in target titles is one of the more defensible outcome measures PR has, especially when the branding team tracks the same measure across other channels.
Referral traffic, links and search effects
Online coverage affects the organisation's site in three visible ways. Referral traffic arrives when readers click a link in coverage. Analytics tools report it by referring domain; the Google Analytics help centre documents how traffic sources and referrals are reported. Filter referral reports by the domains in the coverage log for the relevant period. Where you control the link, such as in a bylined article, a guest piece or an asset you supply, add campaign parameters so the visit is attributed to the campaign; the naming conventions and UTMs guide sets out how to tag links consistently with a UTMURL tags identifying traffic source. scheme. Journalists will not usually add tags to their own links, so referral data remains the main source.
Links from coverage strengthen the organisation's site in search. Log whether each piece links and to which page, and check with link tools that the BacklinkA link from another site to yours. is live and crawlable. A link from a relevant title to a useful page is worth noting in the report; a count of links from irrelevant sites is not.
Search effects show up mostly as branded search: more people searching for the organisation, its products or its spokespeople after coverage. Search console data and analytics show branded queries and organic visits over time. Read them against the campaign calendar and seasonality, and be cautious about claiming causation from a single spike, since other activity may coincide.
Building the coverage report
A coverage report should answer, in order: did we reach the people we needed to, did anything change, and what should we do next. That ordering puts outcomes before outputs and ends with decisions, which is what readers outside PR want. The measurement plan framework is the place where those outcomes and their sources are agreed; the report reflects them rather than inventing new ones each month.
A useful structure has five parts. A summary of a few sentences with the outcome KPIs against target and the main reason for any change. Outcome measures as trends: earned share of voice, branded search, referral visits, and any survey results. Output highlights: coverage in top-tier titles, message inclusion and the best pieces, with links. Competitor and issue notes: what rivals earned and what the conversation is about. And next steps: what to repeat, what to stop, what opportunities are coming. The builder below lets you assemble a report structure from these parts.
Assemble a coverage report
Keep raw coverage lists in an appendix or a linked log. Readers who want to see every piece can; readers who want to know whether PR is working are not made to wade through them. Report total ReachUnique people who saw an ad. with care, if at all: title audience figures overstate who saw a given piece, so present them as context rather than as a KPI.
A monthly coverage report that leadership reads
Pewterledge Bank, a regional lender to small businesses, replaces a 30-page monthly clippings pack with a two-page report. Page one opens with three sentences: earned share of voice in 10 target business titles was 18% against a target range of 15% to 20%, branded search was 9% above the same month last year, and referral visits from coverage totalled 1,240, mostly from one feature on cash-flow lending. Page two lists the 6 tier-one pieces with links, notes that a competitor's rate announcement dominated the trade press, and proposes a data story for next month on late payments. The full log of 41 pieces sits in a linked spreadsheet. The executive team starts asking questions about the report, which it had never done with the clippings pack.
Cadence and who reads it
Different readers need different reports at different speeds. The PR team needs near-real-time monitoring to respond to coverage, correct errors and spot opportunities, usually a daily digest of alerts. The marketing leadership needs a monthly report on outcomes and highlights, aligned with other channel reports so that PR's measures sit beside paid, search and social. The executive team or board needs a quarterly view focused on outcome trends, competitors and any impact evidence, usually as a short section of a wider marketing report rather than a separate document.
During a crisis or major launch the cadence speeds up: daily summaries for the people managing the situation, with the monthly report capturing the aftermath. After each major campaign, a short review comparing what ran with what was planned feeds the next round of planning.
Whatever the cadence, consistency matters more than polish. The same measures, defined the same way, reported on the same schedule, build the trend lines that let readers judge whether PR is working. The PR measurement hub sets out how those measures connect to the wider plan.
For an individual release, the press release distribution checklist makes setting up monitoring and agreeing the outcomes to judge part of the pre-send checks, so the first coverage is counted.
Common mistakes
- Monitoring without a scope. Unbounded queries create noise; define brands, people, competitors and topics.
- Logging only good coverage. A log that omits negative or weak pieces cannot explain results.
- Changing the competitor set. Share of voice is only comparable over time with a fixed set of rivals and titles.
- Reporting raw reach totals as a KPI. Title audiences are not the people who saw a piece.
- Clippings packs instead of reports. Lists of coverage are an appendix, not the report.
- Ignoring referral and search data. The web effects of coverage are among the few things PR can observe directly.
- Claiming every search spike. Check the campaign calendar and seasonality before attributing a change to coverage.
- A different report every month. Consistent measures and structure are what make trends readable.
