Data-Led Digital PR for Health and Nutrition Desks

The work starts in a public dataset, not in a pitch list. NHANES interviews and examines about 5,000 people a year and publishes the microdata; USDA FoodData Central holds the composition records behind most nutrition reporting; CDC WONDER, MEPS, FAERS and ClinicalTrials.gov are all queryable and all under-read by the health desks that would happily cover them. A health or food journalist links to a source with a method attached. That is the whole mechanism, and everything below is how we run it. The asset: One study per cycle, not a scatter of ideas. We take a dataset, ask a question of it that a reporter can put in a headline without distorting it, and publish the analysis on your domain with the query, the sample, the years covered, the exclusions and the limitations stated in the open. The limitations section is not a liability, it is the reason a science desk trusts the piece. Where the finding needs a clinician or a registered dietitian to interpret it, we get one on the record with a name. Who we pitch: Named reporters on health, nutrition and food policy desks, plus the trade press that the national desks read before they write. Pitches go out individually with the data appendix attached, under embargo where the story warrants it. We do not send a release to a bought media list, and we do not follow up four times. What we refuse to pitch: Surveys of two thousand adults dressed up as evidence in a category where people's decisions have clinical consequences. Anything riding a current health scare, an outbreak, a recall or a drug shortage. Rankings of the worst foods or the unhealthiest cities, which get coverage and cost you the medical audience permanently. Anything your regulatory reviewers have not seen. And any story that only works if the reader does not check the sample size. What comes back, honestly: Earned links from a study of this kind are unevenly distributed. A piece picked up by one national outlet and its syndication chain can produce most of the referring domains for the quarter, and a piece that lands nowhere still produces an asset your sales team uses. We report both outcomes the same way. If two consecutive studies fail to place, we will tell you the problem is the subject matter or the domain's own credibility rather than the outreach, and we will say which. Measurement: Referring domains with the date, the outlet, whether the link is followed, and whether the coverage names your brand or only your data. Syndicated duplicates counted once. We separate first-party coverage from aggregator reprints in the report, because the second category inflates every digital PR dashboard in this industry and it should not inflate yours. Not for you if: You need coverage tied to a product launch date. Data journalism does not schedule around your campaign calendar, and a study written to support a launch reads like one. Also wrong if your legal team requires approval of a journalist's angle before an interview, which no newsroom will accept, or if the domain publishing the study has a thin or contested reputation in the category, in which case the fix is the site rather than the outreach. Working arrangement: This runs as a retainer because the second and third cycles are worth more than the first: reporters who have used your data once open the next email. Each cycle covers subject selection, the analysis, the write-up and visualisation, the outreach window, and a written post-mortem naming what did not work. Your team supplies the reviewers and the named expert. We supply everything else, including the refusals.