---
title: "How we build data-led digital PR campaigns that earn coverage"
url: https://marketingsignals.com/insights/data-led-digital-pr-campaigns
type: Article
date: 2026-11-01
author: "Jade Denby"
---

# How we build data-led digital PR campaigns that earn coverage

How we build data-led digital PR campaigns: data journalists cannot get elsewhere, tested angles, regional cuts and every link tracked to its target page.

## Key takeaways

- Choose the commercial pages a campaign must support before choosing the idea, so every link has somewhere useful to send value.
- The strongest data is data a journalist cannot get elsewhere: first-party data, public datasets combined in a new way, or Freedom of Information responses.
- Draft and test headlines before commissioning research, and drop any angle that would not survive a news editor.
- Regional and sector cuts of one dataset often win more linked coverage than the national pitch, as long as each sub-sample holds up.
- Report linked placements that reach the target pages, and what happens to those pages' rankings and revenue, not a raw coverage count.

Most digital PR campaigns that fail do so before the first email goes out. The data is a survey any brand could have commissioned, the headline is the one the brand wanted rather than the one a journalist needs, and the links that do arrive point at the homepage instead of the pages that sell.

This article sets out how we take a campaign from a spreadsheet to national and regional coverage: where we find data journalists cannot get elsewhere, how we test angles before pitching, why we cut every dataset for local press, and how we track each link back to the page it was meant to support.

It should help you judge a campaign proposal before you approve the budget, and know what to ask for when the results come in.

## What a campaign is worth to the business

A data-led campaign is one of the larger single investments in a search budget. Research, design and outreach are paid for before a single piece of coverage lands, so a campaign that misses leaves little behind but the invoice. Leadership sees the cost; it rarely sees why the story did not run.

When it works, the return is different in kind from other [link building](https://marketingsignals.com/service/saas-link-building/link-building-services). [Data-led PR](https://marketingsignals.com/glossary/data-led-pr) gives journalists a reason to cite you that has nothing to do with your product, which is why it can win national and regional titles a straight pitch would not. Google says it [uses links as a signal](https://developers.google.com/search/docs/crawling-indexing/links-crawlable) when determining the relevancy of pages, and an editorial link from an established publication is one a competitor cannot simply buy.

That distinction matters. Google's [spam policies](https://developers.google.com/search/docs/essentials/spam-policies) list buying links, exchanging goods or services for links, and advertorials that pass ranking credit as link spam. A campaign that earns coverage on merit carries none of that risk, and the [earned media](https://marketingsignals.com/glossary/earned-media) it produces keeps working in brand searches, sales conversations and AI answers long after the outreach stops.

The commercial question is where those links point. Coverage aimed at pages that already sit on page two for profitable terms can move revenue. Coverage that lands on a homepage, or on a campaign page with no route to anything you sell, mostly moves a coverage report.

## How we run a data-led campaign, step by step

Every campaign we run follows the same eight steps. The order matters: the first two decisions are made before anyone pitches an idea, and they decide whether the idea is worth having.

1. Start with the pages that need authority. Before any ideation we list the commercial pages sitting between positions 5 and 20 for valuable terms and compare their referring domains with the pages that outrank them. The campaign topic has to sit close enough to one of those pages that a reader would happily follow a link from the data to it.
2. Find data journalists cannot get elsewhere. We look first at the brand's own aggregated, anonymised data, then at public datasets combined in a way nobody has published (the ONS release calendar shows what official statistics are due and when), then at Freedom of Information requests, which public authorities must normally answer within [20 working days](https://ico.org.uk/for-organisations/foi/freedom-of-information-and-environmental-information-regulations/time-limits-for-compliance-under-the-freedom-of-information-act-section-10/). A commissioned survey comes last, because newsdesks receive a great many of them and a poll rarely tells a journalist something they could not have guessed.
3. Write the headlines before the research. We draft the five to ten headlines the data would need to produce for a national, a regional and a trade title to run it. If none of them would survive a news editor, we change the research rather than hoping, which also stops anyone fishing through the results afterwards for something to pitch.
4. Test the angles before pitching. We check each headline against what has been covered in the past year, look at search interest in the underlying topic, and quietly sound out a few journalists we know on the two strongest angles. We lead with the angle that tests best, not the one the brand likes best, and keep the rest as follow-up hooks.
5. Check the numbers hold up. We hold every survey to the [British Polling Council's disclosure rules](https://www.britishpollingcouncil.org/rules-of-disclosure/), which require members to publish the client, fieldwork dates, method, sample size and geographic coverage, with full tables available within two working days. If a regional or demographic sub-group is too small to quote with confidence, that cut does not go out.
6. Cut the data for regional and sector press. One dataset becomes a national story, a set of city or regional stories with local figures and a local comparison, and one or two trade stories for the sector the brand sells into. Regional titles run stories about their own area that they would never run about the country as a whole, so the cuts often produce more linked coverage than the national pitch.
7. Host the data on the main site, next to the page it supports. Findings, a table or map, the methodology, a downloadable file and a spokesperson quote sit on one page on the brand's own domain, with a contextual internal link to the commercial page the campaign was built for. Journalists link to a source they can cite; they rarely link to a homepage.
8. Pitch in waves, then track every link back to its target. We go to nationals first under embargo, regionals next with their own figures, then trade, following the principles in our guide to [pitching a journalist](https://marketingsignals.com/insights/how-to-pitch-a-journalist-what-ex-reach-plc-pro-thinks-of-your-pitch). Every placement is logged with its URL, whether it links, where the link points and its rel attribute, and unlinked mentions get a polite request for a source link, because a link to the data page is the one that passes value on to the page that sells.

## How to measure it

We report a campaign against the page it was built to support, not against a coverage target. The coverage log is the base layer and the commercial measures sit on top of it.

- Placements by tier (national, regional, trade), with the share that carry a followed link.
- Links reaching the data page or the target page, and new referring domains to each, compared with the gap identified in step one.
- Rankings and organic clicks for the target page's main terms, read over three to six months rather than the week after launch.
- Referral sessions from coverage, which GA4 shows by session source and medium in the [traffic acquisition report](https://support.google.com/analytics/answer/12923437).
- Branded search demand, and any direct approaches from journalists who saw the story.

## What to report upwards

Search Console's Links report is a useful cross-check on which pages are gaining links, but Google describes it as not a comprehensive list and caps its tables at 1,000 rows, so we treat it as confirmation rather than the record.

For leadership the useful summary is short: what the campaign cost, how many linked placements reached the target pages, and what has happened to rankings, traffic and revenue on those pages since. Good looks like links concentrated on the pages that matter and those pages climbing for the terms they were chosen for.

## When a data-led campaign does not fit

Data-led campaigns are not the answer to every authority problem. They are a poor fit when:

- The site has technical or content problems that would stop a stronger page ranking anyway. Fix those first.
- The budget covers one campaign a year. A single campaign produces a spike; a steady reactive programme often does more for the same money.
- Every claim needs regulatory sign-off that takes weeks, which drains the news value from most data.
- No commercial page sits close enough to the story to benefit from the links.

## The mistakes we see most

- A nationally representative survey with an obvious answer, pitched as news.
- Data that exists only to flatter the product, which journalists spot at once.
- Regional figures quoted from a handful of respondents.
- Hosting the data on a microsite, a separate domain or a PDF, with nowhere for value to flow.
- Counting pieces of coverage instead of links to the pages that matter.

## Questions to ask your team or agency

Whether the campaign is run in-house or by a specialist [digital PR team](https://marketingsignals.com/service/digital-pr) like ours, these questions separate a plan built for coverage from one built for commercial results.

- Which commercial pages is this campaign meant to support, and how will a journalist's link reach them?
- Could a journalist get this data anywhere else, and if so, why would they use ours?
- Which angles did you test, and what did journalists say before the full pitch went out?
- How many regional and trade cuts will the sample support without stretching it?
- How will you report links to the target pages separately from total coverage?

We run data-led campaigns for [B2B](https://marketingsignals.com/service/digital-pr/b2b-pr), [tech](https://marketingsignals.com/service/digital-pr/tech-pr), [fintech](https://marketingsignals.com/service/digital-pr/fintech-pr) and [fashion and retail](https://marketingsignals.com/service/digital-pr/fashion-retail-pr) brands, with the data and angles chosen for each sector’s press.

## Frequently asked questions

### What is a data-led digital PR campaign?

A data-led digital PR campaign uses original or newly combined data to create a story journalists want to report, such as a regional ranking, a trend analysis or Freedom of Information findings. The data sits on the brand's own site with a methodology, so publications that cover it link back to it as the source. The aim is editorial links and coverage that support specific commercial pages, not only awareness.

### How long does a digital PR campaign take?

Plan for several weeks of preparation before launch. Sourcing and analysing data, testing angles, building the asset and preparing regional cuts all take time, and Freedom of Information requests alone can take 20 working days to come back. Most coverage arrives in the days after launch, with a tail from follow-up angles. The effect on rankings for the pages a campaign supports usually shows over three to six months.

### What makes a good digital PR example?

Good digital PR examples share three traits: data a journalist could not get elsewhere, a headline that works without the brand's name in it, and versions of the story for regional or trade press. When reviewing an agency's examples, look past the logos on the coverage list and ask where the links point. A strong example sends links to a page on the brand's own site that supports something it sells.

## Sources

- [Spam policies for Google web search: link spam, Google Search Central](https://developers.google.com/search/docs/essentials/spam-policies)
- [Link best practices for Google, Google Search Central](https://developers.google.com/search/docs/crawling-indexing/links-crawlable)
- [Time limits for compliance under the Freedom of Information Act (section 10), ICO](https://ico.org.uk/for-organisations/foi/freedom-of-information-and-environmental-information-regulations/time-limits-for-compliance-under-the-freedom-of-information-act-section-10/)
- [Release calendar, Office for National Statistics](https://www.ons.gov.uk/releasecalendar)
- [Overview of the rules of disclosure, British Polling Council](https://www.britishpollingcouncil.org/rules-of-disclosure/)
- [[GA4] Traffic acquisition report, Google Analytics Help](https://support.google.com/analytics/answer/12923437)
- [Links report, Search Console Help](https://support.google.com/webmasters/answer/9049606)

Visibility Playbooks

## Related Playbooks

More Insights

## More articles we think you’ll like