How to set digital PR service standards for data and sources

What to take away

Digital PR service standards are the written rules that decide which data a campaign may cite, how the numbers get checked, and who signs them off.

Why source rules belong in the service standard

A service standard is a promise about process. It tells a client what your team will always do, and what it will refuse to publish. Source rules are the part clients test hardest, because one weak statistic can undo a campaign.

Deciding what counts as a primary source

Name the bodies you accept: the Office for National Statistics, regulators, peer-reviewed journals and a client's own audited accounts. Everything else is secondary and needs a visible label.

Audience measurement shows why this matters. A panel figure and a census figure are both numbers, but they support different claims. Nielsen's review of big data in audience measurement sets out what large datasets add and where coverage gaps appear. The method note therefore travels with the figure.

Recording method and sample size

For every figure, log the fieldwork dates, the population, the sample size and the exact question wording. A survey of 200 respondents cannot carry the same headline as one of 2,000. If a source does not publish its method, the number stays out of outreach copy.

Source checks also need a home in the wider delivery cycle. Our parent guide to digital PR operations and delivery shows where they sit in the weekly workflow, from planning to reporting.

Writing standards a team can follow

Turning the rules into a checklist

A standard nobody can follow at speed is decoration. Convert it into five lines that an account executive completes before a figure leaves the building.

Approval and version control

Give one named person the authority to approve data-led copy, and a second reader for any headline statistic. Keep the approved source list in one place, with a review date beside each entry. When a figure is superseded, mark it retired rather than deleting it, so nobody revives it by accident.

Checking the evidence before publication

Spotting misleading claims

A claim can be accurate and still mislead, if the sample or the comparison is chosen to flatter. The Consumer Protection from Unfair Trading Regulations 2008 deal with misleading commercial practices. The explanatory memorandum to those regulations sets out the background and the tests applied.

Where a business advertises to consumers, persistent problems can be referred to Trading Standards, as the ASA explains on its page about Trading Standards referrals. Our standard therefore asks for a second reader before publication.

When a claim fails review

Pull the figure first, then decide whether to replace it or drop the sentence. Log the failure with its cause, because patterns show which sources keep letting the team down. Tell the client before they read it elsewhere.

The pass or fail test is short enough to memorise. A fuller version sits in our data and sources quality checklist for teams that want a written audit trail.

Common questions

Which figures need a method note?

Any number that appears in outreach copy, a client report or a pitch deck. If a figure is used internally to guide planning, a short note is enough.

Who should sign off a data-led campaign?

One named owner, usually the account director, plus a second reader for the headline statistic. That owner should explain the method without opening the file.

How often should the source list be reviewed?

Quarterly works for most agencies. Add an immediate review whenever a source is corrected, withdrawn or found to have changed its methodology.

What do we do when the client supplies the data?

Hold it to the same test. Ask for the raw file, the method note and the fieldwork dates, and record who confirmed them.