Introduction
Every public relations effort ends in a question: did it work, and what did the world actually say? Press tracking is the discipline that answers it. Stripped of tool marketing, press tracking is the systematic capture and careful reading of your coverage — where it ran, what it said, what it got right, what it got wrong — done continuously rather than in a panic when someone senior asks. It is the feedback loop of the whole PR function: the pitching earns the coverage, but the tracking is what turns coverage into knowledge the next decision can use.
Without it, distribution is broadcasting into the dark. Releases go out, interviews happen, and nobody can say with confidence what appeared, where, in what form, or with what effect — only that activity occurred. With it, done honestly, an organisation knows its own public record: the accurate stories and the errors, the themes gathering strength and the ones fading, the journalists who cover it well and the questions that keep recurring. This guide treats press tracking as that discipline — a craft of definitions, search terms, sceptical reading and usable reporting — rather than as a software purchase.
What Counts as Coverage
The first act of honest tracking is definitional, because at least three different things travel under the word “coverage” and they are not worth the same. The first is earned editorial: a journalist or outlet chose, independently, to write about you — a story, a review, an interview, a mention inside a wider piece. Someone with no obligation to you decided you were worth their readers’ attention. This is the coverage PR exists to earn, and the only kind that carries third-party credibility.
The second is your own release, republished. Distribution and syndication feeds carry releases verbatim to dozens of sites, where they appear dressed as articles — your headline, your quotes, your claims, unexamined. It has legitimate uses as a record and a footprint, but editorially it is you talking about yourself. The third is paid placement: advertorials and sponsored content, bought and labelled (or required to be labelled) as advertising. Also legitimate, also not earned.
Tracking reports routinely blur the three into one impressive total, and the blurring is where the discipline fails. A report that counts your own syndicated release as “coverage” dozens of times over is measuring its own echo and calling it applause. The honest tracker separates the categories every time — earned, syndicated, paid — and reports them separately, because each means something different and only one of them is evidence that independent voices found you worth covering. Definitions first; totals afterwards, if at all.
Building the Watch: Search Terms Are the Craft
A tracking watch is only as good as its search terms, and building them is the craft at the centre of press tracking. Start with the obvious and then do the unglamorous work the obvious skips: the company name in its full, shortened and commonly misspelt forms — journalists mistype, subeditors abbreviate, and a watch set to one exact spelling silently misses real coverage. Add product and service names, current and recent. Add named spokespeople, because coverage often quotes the person without leading with the brand. Add campaign phrases, report titles, event names — the distinctive strings by which your own stories can be recognised in the wild.
Calibration is a balance with failure on both sides. Too narrow, and the watch misses the story that mattered — the piece that used the chief executive’s name and not the company’s, the review under a product nickname. Too broad, and it drowns its reader in passing mentions of common words, until the daily digest goes unread and the one important item is buried on page nine of it. The terms need revisiting whenever the organisation changes — new products, new people, new campaigns — because a watch is a model of what you are called and what you are doing, and both drift.
As for capture itself, there are three broad categories of method, and each misses something. Automated alert services are fast and tireless but context-blind; they find the string, not the significance. Monitoring databases and clipping services cover more ground, including outlets alerts never reach, at the cost of money and a time lag. Human clipping judgement — a person actually searching and reading — catches nuance, misattribution and the story that never used your name at all, but cannot match machine volume. Every automated system misses context; every manual one misses volume. The professional answer is a defined combination of the three, written down so it survives staff changes — and tested periodically against a manual search, because the only way to learn what your watch is missing is to go and look for yourself, on a schedule, and compare.
Reading What You Captured
Capture is the easy half. The discipline lives in the reading, and the first thing a reader assesses is prominence. A story built around you — your news in the headline, your voice quoted, your facts carrying the piece — is a different event from a substantive mention inside someone else’s story, and both differ from a passing reference in a list. Log them differently. A tracking report that treats a one-line mention and a full feature as equal units has discarded the most useful information it collected.
The second assessment is accuracy. Coverage contains errors: wrong figures, outdated titles, a product described as doing what it does not do. Errors found are worth correcting — through the outlet, factually, once, with the correct information supplied and no argument about tone or fairness attached. A correction requested cleanly is usually made cleanly, and the corrected record is what your archive keeps. What the reader should not do is litigate opinion. A reviewer’s judgement is theirs; a misstated fact is a fixable thing. Knowing the difference, and acting only on the second, is professional reading.
Read the quote in its context, too — the sentence around yours may reverse its apparent meaning — and treat every audience figure with open scepticism. The reach numbers attached to outlets are, in the end, the outlet’s own marketing numbers: potential audience, not actual readers of your piece, assembled by whoever benefits from them being large. Use them, if at all, as rough orders of magnitude. And treat with particular suspicion any conversion of coverage into an equivalent advertising cost — the practice of pricing editorial as if it were bought space. It flatters every report it touches, it compares things that are not comparable, and the measurement profession itself has long since disowned it as a vanity metric. A tracker who reports it is decorating, not measuring.
Sentiment, Honestly
Most tracking systems will offer to score each item’s sentiment automatically, and the honest verdict on that offer is: useful as a sieve, worthless as a judgement. Automated scoring reads words, not meaning. It misreads irony and understatement, which are the native registers of much journalism. It stumbles on industry language, where words that sound negative to a general dictionary — disruption, exposure, challenge, aggressive pricing — are neutral or even favourable in context. And it flattens mixed pieces, which is to say most serious coverage, into a single positive-or-negative stamp that describes none of them.
Used properly, the score is a sorting aid: it can push the apparently hostile items to the top of the reading pile so a human sees them first, and it can flag volume shifts worth investigating. The moment it becomes the reported verdict — a sentiment percentage presented to leadership as the state of the organisation’s reputation — it has exceeded what it can bear. For anything that matters, a person reads the piece. Sentiment in press tracking is a triage label, never a conclusion.
Competitor and Issue Tracking
A watch pointed only at your own name tells you what was said about you and nothing about the conversation you were said in. Extend it to competitors, with the same terms discipline applied to their names and products, and the coverage becomes comparative: who is being written about, on what themes, with what prominence. The usual summary of this — share of voice, your proportion of the total coverage across the set — is a rough instrument and should be handled as one. It is comparable only when collected identically: same sources, same terms logic, same prominence rules, period against period. Collected loosely, it produces precise-looking numbers that mean very little, and a change in the watch can masquerade as a change in the world.
More valuable still is tracking the issue rather than the brand. Every sector has live questions — a technology’s safety, a practice’s ethics, a cost, a regulation in draft — that move through coverage in patterns before they arrive at any one company’s door. A watch on the issue’s own terms catches the pattern forming: the commentators aligning, the questions hardening, the story template being built that a future event will be poured into. That is early warning, and it is the highest use of press tracking — the developing crisis visible in coverage patterns while there is still time to prepare an honest answer to it.
Reporting Upward
Tracking earns its keep in the report, and a good report is one executives can act on. Its skeleton is four questions. What ran — the significant coverage, separated into earned, syndicated and paid, with prominence noted. What it said — the substance in a sentence or two each, including the parts that were wrong or unflattering; a report that only carries good news is a liability dressed as a summary. What changed — against last period: the themes rising or fading, the outlets warming or cooling, the issue moving. And what needs action — a correction to request, a claim to substantiate better next time, a journalist to thank, a brief to offer, a question the organisation keeps failing to answer and should prepare for.
The failure mode is the vanity total: a swollen count of mentions, nobody having read them, presented as achievement. Totals without reading conceal everything that matters — that the mentions were your own release echoed back, that the one substantive piece contained an error now lodged in the record, that sentiment moved against you inside the noise. Trends over time beat single-campaign spikes for the same reason: a spike says something happened; a trend says something is happening. Report the trend, attach the reading, end with the actions. That is a document a leadership team can use, and the only kind worth the tracking that produced it.
Cadence and Retention
How often the watch is read should follow how fast the world is moving around you. During a live campaign, a launch or a running story, reading is daily — errors and angles compound quickly, and a correction or a brief offered on day one is worth ten on day five. In ordinary times, a weekly digest and a monthly read of the whole picture serve better than a daily drip nobody opens. The cadence matters less than its regularity: tracking that happens in bursts, when someone remembers, misses exactly the slow movements — the issue gathering, the tone shifting — that continuous watching exists to catch.
Keep what you capture. An organised archive of coverage — the item, the outlet, the date, the prominence, the category, the correction if one was made — becomes the raw material of everything the PR function later has to prove or reconstruct: the periodic report, the campaign review, the audit of what the public record actually says. Retention is unglamorous and decisive. The organisation that keeps its coverage records can answer questions about its own history in minutes; the one that tracked in passing and kept nothing discovers, at the worst moment, that its memory of its own coverage is a pile of other people’s screenshots.
A Worked Example
An illustration, with a fictional company. Harbourlight Foods, a mid-sized producer selling through regional supermarkets, has been doing PR on instinct — releases sent, occasional coverage noticed by accident — and decides to build its first proper watch. The communications lead starts with terms: the company name and the two misspellings of it she has genuinely seen in print, the four product line names, the founder’s name (he is the usual spokesperson, and his surname is the most-misspelt string of all), and the phrase from the current campaign. To the same watch she adds the sector’s live issue — packaging waste — in its own terms, and the two competitors whose coverage she most needs to understand.
For capture she combines free automated alerts for speed with a weekly session of her own searching and reading — the defined combination, written on one page so it survives her holidays. The first month teaches the expected lessons. The alerts over-deliver on a product name shared with an unrelated business, so a qualifier is added. A trade piece misstates a product’s ingredients; she requests a correction factually, once, and it is made. And the issue watch shows the packaging debate moving from campaign-group newsletters into regional papers — before any journalist has connected it to Harbourlight. The company uses the warning properly: it gets its own packaging facts straight, prepares its answers, and briefs its founder, so that when the question arrives it is a question they have already answered honestly to themselves.
Her monthly report fits one page: what ran, in the three categories, prominence noted; what it said, including the error and its correction; what changed — the issue’s temperature, one competitor’s new theme; and three actions, each with an owner. Nothing in it is a vanity total. Within two reporting cycles, the leadership meeting is using it to decide things, which is how she knows the watch is working.
Common Mistakes
Press tracking fails in predictable ways. The recurring errors:
- Counting your own echo. Syndicated copies of your own release logged as earned coverage inflate the record and mislead everyone who reads it, starting with you.
- Trusting automated sentiment. A score that cannot hear irony cannot judge coverage. Use it to sort the reading pile, never as the verdict in the report.
- Tracking only your own name. A brand-only watch misses the competitor context and, worse, the issue moving toward you while it is still someone else’s story.
- Reporting totals nobody can act on. A count of mentions with no reading behind it produces no decisions — and quietly hides the errors and shifts that decisions were needed on.
- Never testing the watch. A watch never checked against a manual search has an unknown miss rate, and its misses are, by construction, the stories you never knew existed.
Conclusion
Press tracking, done as a discipline, is modest in its tools and strict in its habits: define coverage honestly and never count your own echo; build search terms like they matter, because they are the whole watch; read for prominence, accuracy and context rather than totals; treat scores and reach figures as the rough instruments they are; watch the issue as faithfully as the brand; and report what ran, what it said, what changed and what to do — on a rhythm, with the records kept. An organisation that tracks this way knows its own public record. One that does not is simply hoping the record is kind, which is not a measurement strategy at all.
