Digital PR campaigns built on data your company already owns

A campaign built on data you already hold lands coverage in 6 to 14 outlets, of which 4 to 9 carry a followed link. That is the spread across 31 campaigns we ran between January 2022 and June 2026, counting only links a journalist added without being asked twice. The best campaign in that set produced 26 links and the worst produced two. If an agency quotes you a single number with no spread behind it, ask how many campaigns it averages over. What this engagement answers: Whether there is a story inside your own numbers that a newsdesk would run unpaid, and what that story is. Not whether digital PR works in general. One dataset, one decision: publish it or drop it. Who does the work: Two people, and you meet both before anything is signed. Ellen spent six years on a national daily's data desk and builds and checks every figure we publish. Tom ran the investigations bench at a logistics trade title and does the pitching. Both have killed client stories that were already written. Nothing is handed to a junior mailer, and no pitch goes out under a name that does not belong to someone who can answer the reporter's follow-up question the same hour. How the story is built: Days 1 to 5, the dataset goes into pandas and is reconciled against a public benchmark, ONS, Eurostat or your sector regulator depending on what you sell. We are looking for one of two things: a movement of at least 20 percent year on year, or a split that holds by region, so regional desks each get their own line. Days 6 to 14, the finding is tested on three working journalists off the record before a word is written. Days 15 to 35 is writing, charts in Datawrapper so an editor can embed them without booking a designer, and pitching from a list built in Muck Rack and verified by phone that week, because newsroom job moves make a three-month-old list worthless. What happens when the answer is inconvenient: Two versions of this. First, the data is flat. We say so in writing on day 14, you pay the research stage only, EUR 2,400, and we stop. That has happened four times, twice with clients we would have liked to keep. Second, the story is real but unhelpful to you: your category is shrinking, or your own pricing is the outlier. We pitch it as it stands or we end the engagement. We do not soften a finding once a journalist has it, and we will not ask one to. What you get: The cleaned dataset with a methodology note written to survive a fact-checker. The release plus regional variants. The chart pack. The full outreach log, every name, date, reply or silence, yours to keep. A coverage report listing each placement by URL, outlet, publication date and link type, with the reason each declining desk gave. Not included: Paid placements, sponsored slots, and anything on a site with a rate card. When a publication answers a pitch with a price, we tell you which outlet it was and decline; that is not a negotiating position. No link insertions into existing articles. No guest posting. No on-page or technical SEO on your site, and no promise about your rankings, because that outcome depends on a site we do not control. Who this is not for: Companies that need a fixed number of links inside a quarter to satisfy a target. Newsdesks do not work to your board calendar, and we will not buy links to close the gap. Legal or compliance functions that must approve each sentence before it reaches a reporter. Reporters work in hours. Anyone whose only dataset is a 200-response survey run to produce a headline. Editors recognise those on sight, and so do the readers you want. Businesses with no first-party data at all. We can commission a panel study from about EUR 4,000, but say so now, because it moves the timetable by three weeks.