Data-Driven PR: Turning a Number You Already Have into Coverage
Data-driven PR means building a campaign around a number nobody else has, instead of around an announcement. It works because journalists need figures they can attribute, and almost nobody gives them any. A single decent dataset produces more coverage than a year of company news, and the raw material is usually already sitting in a database somebody in your company runs.
Why it works better than announcements
However, a company announcement is interesting to the company. By contrast, a number about the sector interests everyone in it.
There are four practical reasons journalists prefer data:
- First, it gives them something to attribute. After all, a reporter needs a source for the figure in paragraph two. If your figure is the only one available, you are that source.
- It produces multiple angles. So one dataset cut by region, by company size, by month gives ten outlets ten different stories.
- It keeps working. Also, a good data page gets cited for years, by people who never saw the original pitch.
- It cannot be copied. Still, a competitor can match your product claims. However, they cannot match your operational data.
Where data-driven PR data comes from
In order of how much work they take.
First, your own product or operations data. Anonymised and aggregated. Whatever your system records — response times, order sizes, failure rates, seasonal swings. Cheapest and most defensible, because nobody else has it.
A survey you commission. Next, a few hundred respondents in a specific role. Costs money, delivers a headline figure, and is easy for journalists to quote. Also easy to do badly.
Public data nobody has bothered to analyse. Government releases, regulatory filings, planning records. Free, and the work is in the analysis rather than the collection.
Freedom of information requests. Slow, sometimes months, occasionally produces something nobody else has at all.
Making it defensible
This is where most data campaigns fall down, and a journalist who catches it will not come back.
- State the sample. How many records, over what period, from where. “Based on 12,400 service calls between January and August 2026” invites trust; “our research shows” does not.
- Say what it does not cover. If your data is all from one country, or one customer segment, say so in the release. Naming the limit is what makes the rest credible.
- Do not round into a headline you cannot support. If the figure is 31.4 per cent, it is not “nearly a third of all businesses”.
- Distinguish correlation from cause. Two things moving together is a finding. One causing the other is a claim, and usually not one your data can support.
- Publish the methodology on the page. The page with the method on it is the page people cite.
The temptation is always to make the headline stronger than the data. It works once and costs you every journalist who notices.
What a data-driven PR campaign looks like
| Stage | Time | Output |
|---|---|---|
| Choose the question | Few days | One question your sector argues about |
| Pull and clean the data | 1–2 weeks | A dataset with a stated sample |
| Analyse and find the finding | Few days | One headline number, three secondary cuts |
| Build the page | 1 week | Findings, method, charts, downloadable data |
| Pitch | 2–3 weeks | 20–40 journalists, individually |
| Follow-up angles | 2–4 weeks | Regional cuts, reactions, second wave |
Three to five pieces of genuine coverage from one dataset is a good outcome. Anyone promising more is describing paid placement.
Choosing the question
The best data questions share three features: people in the sector disagree about the answer, nobody has measured it, and you happen to be able to.
Practically, look for:
- Something everyone assumes but nobody has checked
- A gap between what people say they do and what your data shows they do
- A change over time that contradicts the accepted story
- A regional or size-based difference nobody has quantified
Avoid questions whose answer is obviously flattering to you. Journalists spot self-serving research instantly, and it is the fastest way to be ignored.
What to avoid
Surveys of your own customers presented as industry research. A biased sample described as a market finding.
Tiny samples with big headlines. Two hundred respondents is fine if you say so. It is not “what British business thinks”.
Data that only supports buying your product. If every finding points at your sales page, it reads as marketing, because it is.
Publishing the chart without the numbers. Journalists need the figures, not a picture of them.
Doing it once. The same study repeated annually becomes a reference point, and the second year is far easier than the first.
Frequently asked questions
What is data-driven PR?
Building campaigns around original data instead of announcements, so journalists have a figure they can attribute and several angles they can use.
How much does a survey cost?
A few hundred respondents in a specific professional role typically runs into the low thousands. Analysing data you already own costs time rather than money and is usually more defensible.
How big does a sample need to be?
Big enough to state honestly. Two hundred is workable if you describe it as two hundred. The failure is not small samples, it is small samples described as large ones.
Do we need a statistician?
For a straightforward count-and-compare, no. For anything claiming a relationship between variables, get someone who can tell you when the data does not support the claim.
How often should we publish data?
Once or twice a year, properly, beats quarterly rushed. An annual study people wait for is worth more than four nobody notices.
Data-driven PR: the short version
Find one question your sector argues about, answer it with data you already have, state the sample and the limits plainly, and pitch it to people who cover the subject. The honesty about limits is not a weakness in the pitch — it is what makes a journalist trust the number.
If you want help turning operational data into something journalists will use, send a brief.
Data-driven PR does not need a research budget: it needs one number nobody else has published.
For the numbers behind this, see press release SEO. Google’s guidance on creating helpful content is the standard this work is ultimately judged against.
That is the whole of data-driven PR, and it is why data-driven PR outperforms opinion pitches.