How to come up with digital PR ideas that journalists cover

A digital PR idea starts from what a journalist's readers care about, then works back to what your brand can credibly say and prove. Generate ideas from five sources, test each against three questions, and write the headline before you spend anything.

In short

  • Practitioners test an idea against three questions, which are whether it is newsworthy without the brand, whether the brand is credible saying it, and whether the data can be gathered in under two weeks.
  • Cision's State of the Media 2026, a survey of 1,899 journalists, found that 82% reject pitches that do not fit their audience or beat, so an idea has to start from a named beat.
  • Journalists in the same survey asked for data or research (47%), embargoed or early access (45%) and access to experts (42%), figures relayed by BuzzStream because Cision's full report is gated.
  • In BuzzStream's State of Digital PR 2026, 73.6% of practitioners said they use AI for ideation, which makes ideas built on data only you hold harder to copy.
  • Speakers at Digitaloft's Digital PR Summit in May 2026 argued that simple ideas beat big ones and that data needs a narrative to land.

Start from the journalist’s reader, not from your brand. A digital PR idea works when a reporter on a named beat would want the story even if your company were not in it, and when your company is a believable source for it. Generate a long list from a few dependable sources, test each idea against three questions, then write the headline and check the data before you spend anything.

This lesson covers finding and choosing the idea. Formats, costs and pitching are on the digital PR tactic page. The idea deserves a lesson to itself because it is the cheapest stage at which to be wrong. An idea dropped at the list stage costs a few minutes, and the same idea dropped after a survey has been fielded and a page built costs the whole campaign, so most of what follows is about finding the weak ones early.

What journalists are looking for

Ideas fail most often on fit. Cision’s State of the Media 2026 surveyed 1,899 journalists. It found that 72% consider a quarter or fewer of the pitches they receive relevant, and 82% reject pitches that do not match their audience or beat. So the first question is always: which journalist, on which beat, would run this? If you cannot name the beat, the idea has no destination, however interesting it seems in the meeting where it was proposed. Those two figures also describe the competition. A reporter who finds most of the inbox irrelevant has every reason to sort on subject before looking at quality.

The same survey says what they want from PRs: data or research (47%), embargoed or early access (45%) and access to experts (42%). Cision’s full report is gated, so those three figures are as relayed in BuzzStream’s 2026 digital PR statistics. It also found that 66% rely on PR-provided content for story ideas, more than any other source. Journalists need ideas. They need ones that fit. That is an encouraging pair of facts to hold together, because it means the door is open to anyone who does the work of matching the story to the beat, and the three things asked for are all things a business can supply: a number, a first look, or a person who knows the subject.

Muck Rack’s State of Journalism 2026, as relayed in secondary summaries we could not check against the original, points the same way: beat relevance was the most valued quality in a pitch at 70%, ahead of interview access at 58% and original data at 40%. With that caveat about the sourcing, two separate surveys pointing at relevance is a good reason to treat it as the first filter and not as one consideration among several.

Five places ideas come from

Source What to look for Example angle
Your own data Sales, searches, bookings, support tickets or usage that show a change or a ranking The most-searched holiday destinations on your site this year against last
Public data Government, statistics office and regulator datasets that can be re-cut by place, age or sector A ranking of cities built from official housing and wage figures
A survey A question people argue about, asked of a representative sample What share of adults could cover a month without pay
The calendar Predictable events such as budgets, tax deadlines, holidays and sports fixtures Expert comment prepared before a budget announcement
The news A running story your expert can explain or add a number to A comment on a new regulation the day it is announced

The examples are illustrations, not campaigns we tracked. The first three sources lead to data-led campaigns. The last two lead to reactive PR. The split matters for planning, because the first group needs time and budget to produce something, while the second needs an expert who is available at short notice and a view prepared in advance. Tools named by practitioners in BuzzStream’s 2026 report and related guides include Google Trends, Glimpse and AnswerThePublic for spotting what people are asking, and national statistics portals such as the ONS and Eurostat for free data.

An idea also needs a shape. BuzzStream’s guide to digital PR formats lists data studies, indexes and rankings of cities or regions, regional maps, recurring annual reports, visual pieces, interactive tools and calculators, expert commentary and product PR. Its examples show how different shapes perform. The link counts are BuzzStream’s, and many of the campaigns predate 2024. As a survey, it gives Preply’s study of subtitle use, with about 800 linking domains and coverage in national newspapers. As an index, it gives Homebuyer.com’s most affordable cities ranking, with more than 80 linking domains, and the page ranks for its main search term. As an interactive test, it gives JustPark’s reaction time game, with more than 1,300 linking domains. And as an original analysis, it gives Backlinko’s voice search study, with more than 3,000 linking domains.

Three things recur. Each one answers a question people already ask. Each gives a journalist a number or a ranking to put in a headline. And each lives on a page worth linking to, which is the same quality that makes a good linkable asset. The spread of results is worth noticing too. These are examples a guide chose to feature, and they still run from more than 80 linking domains to more than 3,000, so a format is a way of giving an idea a usable form and no promise of a particular return.

The three tests

Practitioners commonly run every idea through three questions. This is a working method, not a measured standard. The first question is whether it is newsworthy without the brand. Remove your company name from the headline. If nothing is left, it is an advert. A headline that survives the removal has a subject of its own, such as a ranking or a change, and the brand’s role is to be the source of it.

The second question is whether the brand is credible saying it. A mortgage broker can talk about house prices. It has no standing to rank the best pizza toppings. Credibility also protects relevance: in BuzzStream’s State of Digital PR 2026, 87.2% of practitioners rated site relevance as a criterion for link quality. The two are connected because a story close to what you sell is covered by publications close to your subject, and links from those are the ones practitioners rate. The third question is whether the data can be gathered in under two weeks. If it takes longer, the idea is a research project and the news moment may pass. An idea that fails any one of the three goes back on the list. It is not thrown away, since a different angle or a different data source can sometimes rescue it, but it does not go forward as it stands.

Score the shortlist and write the headline first

When several ideas pass, score them. The grid below is a template of our own, built from the criteria above. Mark each line from 1 to 5.

  • Beat fit: can you name ten journalists who covered this topic recently?
  • Headline: can you write the headline in ten words with one number in it?
  • Regional or sector cuts: can the same data produce a separate story for each city, region or trade?
  • Topical relevance: does the story sit close to the pages you want to rank?
  • Proof: is the data source solid, and can you publish the method?
  • Repeatability: could you run it again next year?

Regional cuts matter because they multiply the number of outlets: one dataset becomes a story for each local paper. Repeatability matters because recurring reports keep earning. BuzzStream cites Bankrate’s annual Emergency Savings Report at more than 8,000 linking domains. The scores are a way of making a group argue about the same things, and the total matters less than the lines where an idea scores low. An idea with a strong headline and no beat fit has a known problem that can be examined, which is more useful than a general feeling that it is risky.

Before any production, write three headlines a journalist might use and the one-line pitch. Then check the data quickly to see whether the headline is likely to be true. If the result is dull, change the angle or drop the idea. This costs an hour and saves a campaign. The order is what protects you. A team that gathers the data first can feel obliged to publish whatever it shows, while a headline written first sets a bar the data has to clear. Speakers at Digitaloft’s Digital PR Summit in May 2026 made two related points. Simple ideas beat big ones. And data without a narrative does not land: the story is the finding and what it means, not the spreadsheet.

Using AI for ideas

In BuzzStream’s State of Digital PR 2026, 73.6% of practitioners said they use AI for ideation. It is useful for widening a list and for finding angles in a dataset. Two cautions apply. An idea that a general tool suggests is available to every other team. And Cision’s 2026 survey found that 53% of journalists object to AI-generated pitches and releases. Use AI to think, and base the idea on something only you have: your data, your expert or your access.

The first caution follows from the adoption figure itself. If close to three quarters of practitioners (the 73.6% above) are asking similar tools similar questions, the ideas those tools offer first may well be offered to several other teams too. That does not make the tools useless. It moves their value to the stages where sameness does no harm, such as listing every calendar event in a sector or suggesting ways to cut a dataset you already hold. The second caution is about the pitch and not the idea, but it belongs here because the two are easily produced together. The three tests still have to be applied by a person, whatever produced the list.

Why ideas fail

Ideas fail in five recognisable ways, and each maps to something above. The first is putting the brand first: the idea exists to mention the product, and no reader would care. The second is having no credible link to the brand, where the story could come from anyone, so journalists credit no one. The third is having no one to send it to. The idea is interesting, but no beat covers it. The fourth is being too big, with months of build for a single moment of coverage. The risk is highest with hero campaigns. The fifth is being off topic, so that it earns coverage that has nothing to do with what the site sells. The first two are caught by the first two tests, the third by the beat-fit line of the grid, the fourth by the two-week test and the fifth by the line on topical relevance, which is why running the tests in full is quicker than it sounds.

Once an idea passes the tests, build the media list before production, because the list tells you whether the audience for the story exists. If you cannot find enough journalists to fill it, you have learned that at the cost of an afternoon. For who can run a campaign for you, see digital PR agencies and tools. Terms used here are defined in the glossary.

Common questions

What makes a digital PR idea newsworthy?

It would still be a story if your brand name were removed. That usually means it is new, it affects a group of readers, it can be compared across places or groups, or it attaches to something already in the news.

Can I use AI to generate digital PR ideas?

Yes, and most practitioners do: 73.6% in BuzzStream's 2026 report. Use it to widen the list, then test every idea yourself, because an idea any tool can produce is one many other teams will pitch too.

How many ideas should I generate before choosing one?

There is no measured benchmark. A common working method is to write a long list of rough ideas, score them, and take the best few to a written headline and a data check before committing budget.

Do digital PR ideas have to be relevant to my product?

They should sit close to what you sell. In BuzzStream's 2026 report, 87.2% of practitioners rated site relevance as a link quality criterion, and off-topic viral campaigns are widely reported to earn links that do little.

What if I have no data of my own?

Use public data, commission a survey, or offer expert commentary instead. A named expert with real credentials is something 42% of journalists asked for in Cision's 2026 survey.

Vendors to look at

  • BazoomEditor's pick

    Sponsored content and link marketplace, managed service

  • MotherlinkEditor's pick

    Backlink services, guest posts, niche edits, full SEO