In The B2B Numbers Census, a website-content audit of 614 B2B companies observed in September 2026 and published in October 2026 by ClaimTheStory Research, 1 in 3 companies never published a number from their own data.
For PR teams, a company’s own data is one place to find evidence a reporter can check. Earning coverage is the goal of data-led PR, never a promised outcome, and a finding only holds up when its source, base sizes and limits are plain on the page. The seven sections below walk through that work in order, from choosing a data route and checking your numbers to shaping press angles and assembling a press-ready package.
What data-led PR is, and what it is not
In data-led PR, a company creates or compiles original data, isolates one finding a journalist’s audience would care about, and offers it with enough methodological detail for the journalist to verify it. Announcing that your company published a white paper is not news on its own; as B2Press’s guide to data-driven PR puts it, “The report is the source. The finding is the news.”
Not news
“Our company published a white paper.”
News
One finding a journalist’s audience would care about, offered with enough method to verify it.
This guide covers developing credible, checkable findings for journalists. It doesn’t cover measuring PR campaign results, choosing or comparing PR agencies, or generic pitching and media-list tactics. ClaimTheStory isn’t a PR agency and doesn’t run outreach campaigns, and nothing in this process guarantees coverage, backlinks, rankings or AI citations.
Where the data can come from
Original data for PR comes from four routes, and none has to hold facts nobody else has to be useful:
- Your own company records: operational, billing or product logs that give evidence about your own customers, though they say nothing about the wider market (see proprietary data for content).
- Public records and datasets: government, regulatory or open databases where you run an original count or combine sources, limited by whatever fields the public agency collects (see how to get data for content).
- Partner data: records held by an association, supplier or complementary company that give a combined view neither party holds alone, subject to written sharing and release terms (also covered in how to get data for content).
- New collection: a structured survey or study built to answer the exact question you want to ask, limited by respondent self-selection and cost (see PR surveys if you plan to commission a survey for press use).
Don’t treat a survey as the default starting point if existing logs or public records already measure the behavior. In Better Agency’s tips for starting a data-led PR campaign, the UK agency describes using UK Freedom of Information requests to power its own campaigns. In the US, the counterparts are federal Freedom of Information Act (FOIA) requests and state public-records laws.
What makes a finding worth a journalist’s time
A reporter needs a finding their audience didn’t already know, about a group their readers care about, that they can verify from your documentation. In Marketing Signals’ guide to data-led digital PR campaigns, the team drafts and tests headlines before commissioning research, and drops any angle a news editor would reject. With your own records, apply the same habit before pulling the extract: write down the headlines you hope to test, and drop any the data can’t support rather than forcing an angle.
The general method for questioning a dataset, running the arithmetic and testing angles is taught in turning data into content. Keep your metric definition exact everywhere it appears: never shorten a defined share into a vague phrase like “tenants stay longer,” and never read a share as a speed, a size or a cause.
Check the finding before anyone pitches it
Run your checks before anyone writes a press pitch. A journalist who spots an unstated exclusion or a tiny regional base drops the piece and may not trust your next one.
For the full method of finding and checking a finding, see turning data into content. If your PR finding comes from a commissioned questionnaire, use PR surveys to check sample and weighting limits.
One dataset, several press angles
Different desks look for different cuts of the same checked dataset: a national business desk wants the overall benchmark, regional reporters look for local numbers, and trade publications want operational context for industry peers. Keep the metric definition and the direction of the action identical across every angle: tenants initiated the move-outs, while facility-initiated ends are excluded and never described as tenants choosing to leave. Never publish a small region or relax your release policy to hand out state or city cuts just to satisfy a local pitch.
Kept
- National business desk: the 24.7% benchmark, with its population stated.
- Regional desks: each region’s figure on its own, once opening dates are checked.
- Trade publications: all four regions with counts and the exact rules.
Rejected
- “Rising rents drive tenants out”: a cause.
- “Facilities pushing out long-term tenants”: reverses who acted.
- “Western tenants move out more slowly”: a share read as a speed.
- “A quarter of America’s storage tenants…”: widens customers to the market.
- “The South is worst at keeping long-term tenants”: a ranking before the check.
The press-ready package: methods note, data page and what stays private
Give journalists and analysts everything they need to inspect your numbers on a single public data page:
- The headline finding with its exact definition and observation period
- The publishable summary table with counts and shares
- A complete methods note (source, period, extraction date, population, definitions, group counts, exclusions, limits, citation line and data-access policy)
- A named press or research contact
Keep your internal working table separate from your public table so suppressed rows and platform-wide totals never slip into print.
No platform-wide total (across all five regions) appears anywhere on the public page. A reader who knew the platform-wide total of 278 facilities or 12,150 move-outs could subtract the four shown regions (275 and 12,000) and recover the suppressed Alaska and Hawaii row (3 facilities and 150 move-outs).
- Platform-wide total 278 facilities, 12,150 move-outs. Never published.
- Four shown regions 275 facilities, 12,000 move-outs. Published.
- The difference 3 facilities, 150 move-outs: exactly the suppressed Alaska and Hawaii row.
Under the team’s illustrative release policy, only region-level shares and counts are published, while facility names, state or city cuts, and tenant-level rows stay private. Always assess disclosure risk with whoever owns the data before publishing; a minimum group threshold alone never makes operational records anonymous. For how to write a full methods note, see writing up survey and research findings, and for fair cross-group comparisons, see how to create a benchmark report.
What earning coverage depends on, and what nobody can promise
Whether a reporter picks up a finding depends on editorial priorities, timing and the news of the day, none of which a content team or agency controls. No honest plan can promise media placements, backlinks or traffic.
You control
A checked finding, its exact definition, a public data page, a methods note and a named contact.
Cannot promise
Media placements, backlinks, traffic.
Before approving any data-led PR project (whether run in-house or by a supplier), ask five questions about it:
- Source: Where do the underlying records come from, and who owns them?
- Metric: What exact unit is counted, and what is excluded?
- Threshold: What minimum group size applies, and is it enforced consistently across all cuts and totals?
- Privacy: What stays private, and has the data owner assessed disclosure risk?
- Sign-off: Who approved the figures and methods note for public release?
Measuring campaign performance after publication is a separate discipline and isn’t covered here. If you are still deciding whether original research fits your budget and goals, start with original research for content marketing. If your project relies on a commissioned questionnaire, check PR surveys before briefing a vendor. Visit our hub on data-driven content marketing to see how all the guides connect.
If you know what you want to say, we find the data that would support it, and tell you if somebody has already published it. Book a demo