Digital PR Measurement Log: Coverage, Mentions, Links

When a digital PR campaign wraps up, the first question is usually "how many links did we get?" That one question squeezes three different events into a single figure: a piece of coverage appearing, your brand being named in it, and the piece actually linking to you. The three can happen independently of each other. A digital PR measurement log therefore keeps them in three columns and never blends the counts. This article explains how to set the log up, which mistakes cause double counting, and why the log should not be treated as proof of results. Akua Adforce is not a PR agency and offers no press release distribution, media monitoring or reporting service. Every number below is invented.

Why a Digital PR Measurement Log Needs Three Separate Columns

One campaign, three different counts

Picture a hypothetical campaign: a data study went out and twelve outlets wrote about it. Eight named the brand, four cited the study as a source, two linked. Saying "twelve results" paints a very different picture from saying "two links." All three counts are correct; they answer different questions. Collapsing them into one figure hides which outcome you were actually pleased with.

Where the log sits in the wider strategy

What digital PR is and how it is run is a separate subject, and the explainer on what digital PR is and how to do it supplies the framework. This article deals only with the measurement log. Rather than leaving measurement until the end of the strategy, decide before the campaign starts which three columns you will keep. Back-filling columns nobody decided on is hard.

Who keeps the log

Give the log a single owner. If two people record the same piece of coverage in different columns, double counting is inevitable. Put the owner's name and the date of the last update in the top row. Counts taken from third-party tools describe what one tool measured and are not Google's own values, and the log should know that from the start.

The Coverage Record: Where, When and in What Format

Fields in the coverage record

The coverage record documents that a piece of content appeared somewhere. Its fields are plain: outlet type, publication date, the address of the piece, its format (news item, interview, column, list) and its relationship to the campaign. Whether the brand was named or a link exists is not this record's job; those belong to the other columns. The coverage record answers a single question: did it run?

Why the address and date are stored

A piece can be deleted, moved or re-addressed. If you keep the full address and the date you first saw it, you can return to the record whenever someone needs proof that it ran. A screenshot alone is not enough; the address and date row is what makes it usable.

Tying coverage to the campaign

Not every piece is a result of the campaign; some run independently. Add a field saying whether a piece is directly tied to the campaign or uncertain, and mark anything you are unsure about as uncertain. For the tracking side, the guide to tracking sponsored posts with UTM parameters and Search Console is a useful companion, though its tracking completes the coverage record rather than replacing it.

The Mention Record: Was the Brand Named, Was a Source Cited?

A mention is not the same as coverage

A piece can run without naming your brand; a general sector story may only say "a study found." The mention record tracks whether the brand or study was named and whether it was cited as a source. Use a simple scale: named, cited as a source, neither.

Not losing the unlinked mention

Pieces that name your brand without linking usually fall onto a separate list. The guide on turning unlinked brand mentions into backlinks covers what can be done with that list. For the log, what matters is not throwing those rows away and showing them separately in the mention column. An unlinked mention is zero in the link column and one row in the mention column.

A hypothetical scenario: a columnist refers to your study but gives no source. If the team deletes the row as a failure, a chance to send a brief thank-you with a suggestion to add the source disappears too. Keeping the row in the mention column makes that follow-up visible, though nobody can say the follow-up will bring a link.

The quality of a mention

Not every mention carries the same weight: being cited with a finding inside a sentence differs from appearing in a list of names. Record quality with a short label: with finding, in a list, in passing. The person applying the label should know it is a judgment call, and labels can change later. Keep this column free of numeric totals.

The Link Record: Linked, Unlinked and Labeled Coverage

Sub-fields of the link column

The link column splits into three sub-fields: is there a link, where does it point, and how is it qualified. The destination matters, whether it is your campaign page or your home page, because a link to the home page and one to the study page read differently.

Link attributes: sponsored, nofollow, ugc

How a link is qualified affects how you read the log. Google's outbound links documentation recommends sponsored for advertisements or paid placements, ugc for user-generated content and nofollow when the other values do not apply. The comparison in the differences between the main link attributes explains them further. Record the attribute as found; "dofollow" is not an official value, only a casual way of saying no qualification is present.

Paid placements stay in the log

A paid placement appears in the same log with its own label. Under Google's spam policies, paid links are not a violation when qualified with nofollow or sponsored. Marking a row "paid" supports both honesty and correct reading; paid and earned items should never merge into one total.

You do not have to write an amount beside paid rows, because a measurement log is not a budget sheet. It is enough to state that the row is paid and how it was qualified. If the qualification changes later, or the publisher removes it, add a dated note.

Reading at Campaign Level, Not Per Placement

The trap of looking at one piece

Judging a campaign by its most talked-about piece is misleading. A few large outlets and many small ones draw quite different campaign profiles. Placement-level detail lives in the log, but the campaign summary is read on a separate row.

The campaign summary row

The summary row gives a count and a spread for each column: how many pieces, how many named the brand, how many linked. Write the denominator beside each count, as in "eight of twelve named the brand." Do not give the share alone, because small samples distort ratios. Add at least a few words of reading to the row, such as "wide but shallow" or "narrow but linked."

Comparing campaigns

When you place two campaigns side by side, make sure the column definitions are the same. If "mention" means a source citation in one and merely a name in the other, the comparison is meaningless. Write the definitions on the first page of the log and do not change them between campaigns.

A hypothetical case: if mention meant only the brand name in the first campaign and a source citation in the second, the second campaign's mention count would naturally look lower. That drop is a difference in definition, not in performance. Put a version number and date on the definition page, and where an old campaign cannot be refilled under a new definition, mark it "not comparable."

Indirect Indicators Such as Brand Searches and Direct Visits

Where indirect indicators belong

Brand-name searches, direct visits and social mentions may shift after coverage runs. They do not belong in the coverage, mention or link columns; they get their own "indirect indicators" section with a source and date row. Their relationship to the coverage cannot be established; they are only read alongside it.

Watching brand searches

You can watch brand searches in Search Console, but PR is not the only possible cause: season, advertising and other channels play a part. The article on growing brand searches with sponsored articles discusses this indicator in another setting. For the log, what matters is measuring brand searches with the same set of queries and the same date range every time.

Keep it as an indicator

Indirect indicators are never proof of success. Titling the section "indicator, not proof" stops readers from jumping to the wrong conclusion.

Logging Mistakes Such as Counting the Same Coverage Twice

Where double counting comes from

The same story can be counted twice when it appears both at its own address and in a social post. Other publishers republishing an agency's story also create extra rows although the content is identical. Add an "original" versus "republished" distinction to the log and show republished rows apart from the total.

Mixing links and mentions

One link can also be a mention and a piece of coverage. Appearing in all three columns is not double counting, because each column answers a different question. The mistake is adding the three together. Do not add a total row; per-column summary rows are enough.

Confusing PR with reputation work

If PR output and reputation management output go into the same table, measurement blurs. The approach in online reputation management with news sites serves a different aim. Add a campaign code column and mark which effort each row belongs to.

A hypothetical situation: a data study and an executive interview run in the same week. If you mix the mentions from both, you cannot later tell which effort produced what. The code column protects that distinction from the start and makes a campaign-level filter easy at month end.

Reading Measurement With a Causality Caveat

The log does not prove results

A rise in brand searches or traffic to a page after a campaign may or may not be a result of the campaign. Other factors may have been at work in the same period. The log documents what happened and not why. State that caveat in one sentence at the top of the log.

How it relates to the ROI question

Return is a separate exercise outside the log. The guide to how to measure backlink ROI takes up that question; this log does not calculate return. To keep the two apart, do not open a money or conversion column in the log.

The language of interpretation

Phrase what comes out of the log with expressions such as "may have helped" or "seen in the same period." Avoid verbs like "delivered" or "brought." When a decision meeting hears this language, it also learns the limits of the table.

The choice is not only a matter of style. A reader who sees "brought" takes it as a promise and expects the same from the next campaign. Probabilistic wording pulls expectations back to what the log actually holds. A hypothetical sentence: "Brand searches rose in the two weeks after publication; the cause cannot be determined from this log."

Recording Failed Attempts

Unanswered and declined pitches

Most pitches sent during a campaign go unanswered or are declined. Recording them shows the next campaign which audience to approach and in what way. A pitch row holds: to whom (as an outlet type), with what angle, when, and what response came. No journalist or outlet names are written.

Reasons for failure

If nothing is known about the reason, write "unknown." Where something is known, note it briefly: timing, topic fit, a request for more data. Do not mix guesses with facts; "no reply" is a fact, "they found it irrelevant" is a guess.

Showing the failure count too

Writing the number of unanswered pitches beside the successful placements reflects the real effort. A hypothetical example: forty pitches, six replies, four pieces. These ratios represent no real market; they simply recall the distance between effort and result.

A failed-pitch record is a learning inventory, not a self-criticism list. Patterns such as which angle drew more replies or what extra data was requested become visible after a few campaigns. One campaign's log proves no pattern, but if three consecutive logs point the same way, it can be treated as an assumption.

Taking the Log to the Strategy Meeting

A one-page summary beforehand

Do not bring the entire log. One page with the three column summaries, the indirect indicators, the failed pitch count and the causality caveat is enough. The detail tabs can be opened when needed.

Three questions for the meeting

Which column fell below our expectation? Which pitch format produced more coverage? What will we do differently in the next campaign? These questions yield learning rather than promises, and the answers are added to the meeting note.

Letting the decision return to the log

A decision made in the meeting should appear as a change in the next version of the log: a new column, a changed definition or a removed field. If the decision never reaches the log, the meeting remains a discussion.

Add a short change note to the head of each version: what was added, what was removed and why. Someone opening the log months later learns from those notes why the table looks the way it does.

Example Template: A Digital PR Measurement Record Table (Hypothetical)

The campaign, outlet types and counts in this template are invented (hypothetical) and show no real campaign. Counts from a third-party tool describe what that tool measured and are not Google's values.

Fill in the log in this order:

1. Enter each piece with its address and date in the coverage column.
2. Add the brand-name and source status to the mention column.
3. Record the link destination and attribute in the link column.
4. Tag republished items and uncertain ties separately.
5. Build the campaign summary row for each column.

Things that stay out of the log:

• A row that adds the three columns into one total.
• Results that cannot be tied to the campaign written down as certain success.
• Journalist or outlet names.
• Any ranking, traffic or revenue expectation.

Frequently Asked Questions

How are digital PR results measured?

With a log of three columns: coverage appearing, the brand being named and a link being given are kept apart. Indirect indicators such as brand searches sit beside them. Results are read at campaign level and with a causality caveat, because the log records what happened and does not prove why.

Why are coverage, mentions and links counted separately?

They answer different questions and can occur independently. Coverage may omit the brand name, and a mention may carry no link. Adding them into one figure hides which outcome you were pleased with. Separate columns prevent double counting and show the true profile of the campaign.

Is an unlinked mention worthless?

There is no reason to call it worthless; it simply shows as zero in the link column and one row in the mention column. What it is good for depends on the campaign's aim. Keeping it on a separate list leaves open the chance to contact the outlet later, though no outcome should be read into that.

How often should the log be updated?

Monthly while the campaign runs and at intervals afterwards is enough for most projects. Pieces can be removed and new mentions can surface, so look again several times after the campaign ends. Refresh the date row at every update and note what changed.

Should failed attempts be recorded?

Yes. Unanswered and declined pitches show how to approach the next audience and reflect the real effort. Without identifying details, keep the outlet type, angle, date and response status. If the reason is unknown, write unknown rather than a guess.

Does a measurement log prove results?

No. The log records what happened, not why. Other factors may have worked in the same period. Present the figures as a profile of the campaign, not as proof of traffic, ranking or revenue, and prefer wording that signals possibility rather than certainty.

Conclusion and Next Step

What does a three-column log change? It swaps "how many links did we get?" for "what did we see, and what do we not know?" Once coverage, mentions and links stand apart, double counting shrinks, and failed pitches become part of the true picture of the campaign.

Should part of your plan involve paying for a placement, think about it as a buyer. Akua Adforce is not a PR agency; it lists guest post listings from publishers. You choose a listing and add it to your basket, payment is made, the seller approves, and a delivery notice follows, with a dispute possible if something is off. The publication decision, timing and link attribute rest with the seller, and no media coverage or outcome should be expected. Record a paid placement in the log with a "paid" label and note whether it was qualified as sponsored or nofollow.