# How to Measure Digital PR: Metrics That Show It Worked

Digital PR · Published 2026-11-19 · Updated 2026-11-19 · RankMonks
Source: https://rankmonks.com/blog/measuring-digital-pr

> Short answer: Measure digital PR in five layers: the activity you carried out, the quality of the coverage it earned, the SEO outcomes that followed, the change in brand demand, and how often AI answers name you. Keep a log of every placement, and leave out advertising value equivalents and raw link counts.

You measure digital PR by tracking what was published, how good it was and what changed afterwards in search, in brand demand and in AI answers. No single number covers all of that, so the honest approach is a small set of metrics arranged in layers.

This guide covers the five layers, what to record about each piece of coverage, which popular metrics to leave out, how long results take to appear and how to structure a report people will read.

## Key takeaways

- Digital PR is best measured in five layers: activity, coverage quality, SEO outcomes, brand demand and AI answers.

- A coverage log that records every placement, link and mention is the foundation, because every other metric is drawn from it.

- Quality matters more than volume, so judge each placement by the relevance of the outlet, where the mention sits and whether it is accurate.

- Advertising value equivalent, raw link counts and a third-party authority score on its own all mislead more than they inform.

- Coverage appears within weeks, but rankings, branded searches and AI answers take months to respond, so report over quarters.

## Which metrics show whether digital PR worked?

The metrics that show whether digital PR worked fall into five layers: activity, coverage quality, SEO outcomes, brand demand and AI answers. Each layer answers a different question, and each helps explain the one above it.

![Pyramid of five measurement layers, from the base up. Layer 1, activity: stories pitched and replies received. Layer 2, coverage quality: relevant outlets, in the body, accurate. Layer 3, SEO outcomes: referring domains, rankings and clicks. Layer 4, brand demand: branded searches and direct visits. Layer 5, AI answers: how often models name and cite you.](/posts/measuring-digital-pr/measurement-layers.svg)

| Layer | The question it answers | What to track |
| --- | --- | --- |
| Activity | Did we do the work? | Stories produced, journalists pitched, replies |
| Coverage quality | Was the coverage any good? | Placements in relevant outlets, links, mentions, accuracy |
| SEO outcomes | Did search respond? | Referring domains to target pages, rankings, organic clicks |
| Brand demand | Do more people look for us? | Branded search impressions, direct visits, referral visits |
| AI answers | Do models name us? | Share of answers, how you are described, sources cited |

**Activity** is the only layer fully in your control. It shows whether a quiet month was caused by a weak story or by too little outreach.

**Coverage quality** is where most of the judgement lies. Ten placements in publications your buyers read are worth more than a hundred on sites nobody visits.

**SEO outcomes** show whether the links did their job. Watch the pages the coverage pointed to and the commercial pages those pages link to. Our guide to [link building](/blog/link-building) explains why referring domains to target pages tell you more than total links.

**Brand demand** shows whether more people know your name. Branded search impressions in Google Search Console are the simplest measure.

**AI answers** show whether the coverage has changed what models say. Track a fixed set of prompts and record how often you are named, a measure covered in our guide to [AI share of voice](/blog/ai-share-of-voice).

## What should you record about each piece of coverage?

Record each piece of coverage as one row in a coverage log, with enough detail to judge its quality later. The log is the source for every figure in the report, so it is worth keeping carefully from the first placement.

Use these columns:

| Column | What goes in it |
| --- | --- |
| Date | When the piece was published |
| Outlet and URL | The publication and the address of the article |
| Type of outlet | National, trade, regional, newsletter, podcast or blog |
| Campaign | Which story or comment earned it |
| Link | Followed link, nofollow link or no link |
| Target page | Which page on your site the link points to |
| Placement | Body text, quote, caption or list |
| Accuracy | Whether your brand and what you do are described correctly |
| Relevance | Whether the outlet's readers are your buyers, scored simply as high, medium or low |
| Notes | The journalist, any follow-up and anything unusual |

Fill it in by hand. Monitoring tools and alerts will find most coverage, but someone has to open each article to judge placement, accuracy and relevance.

Record mentions without links as carefully as links. A named mention in a trusted publication adds to what AI models read about you, as our post on [brand mentions](/blog/brand-mentions-ai-search) explains. It is also a prompt to ask, politely, whether a link to the source could be added.

For example, a workplace safety software company runs one data story in a quarter. Its log shows fourteen placements: two national, seven trade, three regional and two newsletters. Nine link to the research page, three link to the homepage and two are mentions only. Eleven are rated high for relevance. That one table tells the company more than a headline total ever would.

## Which digital PR metrics should you leave out?

Leave out advertising value equivalent, raw link counts and third-party authority scores quoted on their own. All three are easy to produce and easy to inflate, and none of them tells you whether the campaign helped the business.

![Comparison of three metrics to leave out and what to report instead. Leave out advertising value equivalent, a price put on space you did not buy; report instead the relevance of each outlet and whether its readers are your buyers. Leave out raw link counts, where every link counts the same; report instead referring domains to the pages you wanted to lift. Leave out an authority score on its own, a tool's estimate; report instead the placement, meaning where the link or mention sits and what it says.](/posts/measuring-digital-pr/leave-out-report-instead.svg)

**Advertising value equivalent.** Known as AVE, this puts a price on coverage by asking what the same space would have cost as an advert. It measures the size of the space, not whether anyone relevant read it or what the article said. The PR industry's own measurement bodies have advised against it for years.

**Raw link counts.** A total treats a link from a respected trade title and a link from a scraped copy of a press release as equal. Count referring domains instead, and look at which pages they point to.

**Authority scores alone.** Scores such as Domain Rating or Authority Score are estimates made by tool providers, and Google does not use them. They are useful for comparing two sites quickly. They say nothing about whether a site is relevant to your buyers.

**Potential reach.** Figures for "opportunities to see" add up the total audience of every outlet, as if each reader saw each article. Treat them as a ceiling, not a result.

**Syndicated copies.** When one article is republished across a network of sister sites, count it once and note the syndication. Listing every copy as a separate placement inflates the total.

If a report leans heavily on any of these, ask what sits underneath. It is one of the questions worth putting to any supplier, as our guide on [how to choose an SEO agency](/blog/how-to-choose-an-seo-agency) suggests.

## How long does digital PR take to show results?

Digital PR shows coverage within weeks, search results within two or three months and changes in brand demand and AI answers over three to six months. The layers respond at different speeds, so each one needs its own timescale.

![Timeline of when digital PR results appear. Weeks 1 to 4, coverage lands: placements, links and mentions are logged and referral visits arrive. Months 2 to 3, search responds: new links are crawled, target pages start to move and non-brand clicks grow. Months 3 to 6, brand and AI: branded searches rise, models start to name you and your share of answers moves.](/posts/measuring-digital-pr/results-timeline.svg)

**Coverage comes first.** Most placements appear in the first few weeks after a pitch, with a tail of later pieces from writers who find the research through search.

**Search follows.** A new link has to be crawled and weighed with everything else Google knows about the page. Expect gradual movement over a quarter, not a jump the following week.

**Brand and AI come last.** Branded searches rise as more people meet your name. AI answers change as new coverage is retrieved by assistants that search the web, and more slowly as models are retrained on newer text.

These are typical ranges, not promises. A competitive market, a new domain or a thin site will all slow things down. Coverage also cannot rescue pages that have [technical problems](/blog/technical-seo-audit) or do not match [search intent](/blog/search-intent), which is why it works best alongside the foundations in our guide to [SEO in the AI era](/blog/seo-in-the-ai-era).

## How do you report on digital PR?

Report on digital PR once a month in a short written document that covers what was done, what landed, what it changed and what happens next. A report is useful when someone can read it in five minutes and know what to decide.

Build it in this order:

1. **Summary.** Three or four sentences: the main result, the main disappointment and the plan.

2. **Activity.** Stories produced and pitched, and to how many journalists.

3. **Coverage.** The placements from the log, grouped by type of outlet, with the best three shown in full.

4. **Quality.** How many were relevant, linked, in the body and accurate.

5. **SEO outcomes.** New referring domains to target pages, and how those pages are ranking.

6. **Brand and AI.** Branded search impressions and share of answers, compared with the baseline.

7. **What we learned.** Which angle and which outlets responded, and which did not.

8. **Next month.** The stories in production and the dates they depend on.

Three habits keep the report honest.

- **Compare against a baseline.** Record referring domains, branded impressions and share of answers before the first campaign, so later numbers mean something.

- **Report the misses.** A campaign that earned nothing teaches you about the story or the list.

- **Do not claim what you cannot show.** Rankings move for many reasons. Say that coverage contributed, and show the timing, instead of claiming every gain.

Digital PR numbers belong beside the rest of your search reporting, not in a separate file. Our guide to [SEO reporting](/blog/seo-reporting) covers the wider report, and [AI source analysis](/blog/ai-source-analysis) shows how to check whether the outlets that covered you are the ones models cite. If the answer is no, that is your target list for the next [digital PR](/blog/digital-pr) campaign.

## Frequently asked questions

### What is a good number of links for a digital PR campaign?

There is no fixed number of links that makes a digital PR campaign good, because results depend on the story, the market and the outlets. A handful of links from publications your buyers read can do more than dozens from unrelated sites. Judge a campaign by the relevance of what it earned and by what happened to the target pages afterwards.

### Do nofollow links and unlinked mentions count?

Nofollow links and unlinked mentions do count, although they do not pass ranking credit the way a followed link does. They send readers, raise awareness and add to the text that AI models learn from. Log them separately so the report shows the full picture without overstating the SEO effect.

### How do you measure digital PR's effect on AI answers?

You measure digital PR's effect on AI answers by running the same set of buyer questions through the main assistants before and after a campaign. Record how often your brand is named, how it is described and which sources are cited. Our guide to [AI visibility](/blog/ai-visibility) explains how to set up the baseline.

### Can you measure the return on investment of digital PR?

You can estimate the return on investment of digital PR, but you cannot measure it precisely. Coverage contributes to rankings, traffic and enquiries alongside many other things, so attribution is always partial. Track organic enquiries and pipeline from the pages the campaign supported, and present the link as a contribution.

Measurement only works when the reporting is honest about what landed and what did not. Every placement, link and mention is logged in our [digital PR](/services/digital-pr) work, along with what it did for rankings and AI answers. If you want a clearer view of what your coverage is doing, [start a conversation](/contact).
