5W PR Statistics Audit: 46 of 100 Claims Linked Directly to the Original Research or Data Producer

Only 46 of 100 PR statistics in five public roundups linked directly to the organization that produced the research or data. A 5W audit found 31 claims led to secondary pages, 11 to aggregators or report marketplaces, and 12 had no usable source. Readers cannot check a number they cannot trace.

Updated October 27, 2026. An earlier version of this page reported 44 original-source claims and 5 with no usable source. After a coding review, the counts above apply, and every coded entry is now shown in Table 6.

What is the 5W PR Statistics Citation Audit?

The 5W PR Statistics Citation Audit is a claim-level review of five public PR-statistics roundup pages, captured on October 27, 2026. It tested whether a reader could move from a published number to the organization that produced the evidence in one click. It did not test whether each statistic is true.

• What did the audit find?
• How did the five publishers compare?
• Is a link the same as evidence?
• Can readers see every coding decision?
• Why does citation discipline matter for PR?
• How was the audit run?
• What are the limits of the findings?
• How should this audit be cited?
• Frequently asked questions

What did the audit find?

Fewer than half of the 100 audited claims linked directly to the producer of the underlying research or data. The other 54 sent readers somewhere else or nowhere usable.

• 46 claims linked directly to the producer of the survey, dataset, platform analysis or market estimate.
• 31 claims linked to a secondary article, advice page or another statistics roundup.
• 11 claims linked to a data aggregator or report marketplace rather than a clearly identified producing source.
• 12 claims had no usable source: one had no link attached, four had links that were broken, irrelevant or did not show the cited figure, and seven led to a vendor page that did not establish the claim's research or producer.
• 30 claims led to a destination that publicly showed some claim-specific methodological detail, such as sample size, field dates, data scope or method. All 30 were among the 46 original-source claims.

How did the five publishers compare?

Eighty-eight claims led to a destination classified as usable under the audit's rules. The problem was where the link went: 31 stopped at secondary content, 11 stopped at an aggregator or marketplace, and 12 had no usable source.

Performance differed sharply by page. Sprout Social linked 19 of its 20 sampled entries directly to original producers, and BlueTree linked 17. PRLab linked seven, Scopic Studios linked two and flair linked one. The two Scopic Studios entries went to BuzzStream's own digital PR survey page; the rest of its citations led to PRLab, Semrush, Bright Valley or BlueTree pages.

These scores cover only the first 20 eligible entries on each sampled page. They do not represent every statistic on the page or the publisher's work as a whole.

Is a link the same as evidence?

No. A citation can look authoritative while leading only to another article repeating the same number. The audit separated four kinds of destination and recorded figure confirmation as a separate test.

Method visibility was a separate and stricter test. A destination received credit only when the public page disclosed something specific enough to evaluate the claim, such as sample size, field dates, population, data scope or analytical method. Thirty of the 100 claims met that test. The measure is broader than sample size plus field dates, so the 30 does not mean every one of those destinations disclosed both. Sixteen of the 46 original-source claims did not earn method credit, mostly because the producer's page was a landing page or sat behind a paid report.

Figure confirmation was completed for only part of the original-source set. Of the 46 original-source entries, the exact figure was confirmed at the destination for 6, partly confirmed for 1, not confirmed for 11 and not checked for 28. Producers revise their estimates: IBISWorld's current page shows a different 2025 employment figure than the one cited, and the entry still counts as original because IBISWorld produced the estimate.

Can readers see every coding decision?

Yes. Table 6 lists all 100 coded entries from the audit workbook, with the cited source, destination domain, category, method-visibility code, figure confirmation and the reason for the code. Entries marked "not re-checked" keep the original auditor code, and their destinations were not revisited in the coding review.

Why does citation discipline matter for PR?

PR asks journalists and audiences to trust information supplied by organizations with an interest in the outcome, so citation discipline is part of the product. Linking to the original source lets a reader inspect the population, timing and method before repeating the number. Linking to another roundup shifts that work to the reader and raises the chance that context disappears as a statistic travels.

How was the audit run?

Pages were collected on October 27, 2026. The corpus was purposive, not random. Pages were selected because they were publicly accessible, explicitly presented PR or digital-PR statistics, and contained at least 20 quantitative entries. The first 20 eligible entries in body order were kept from each page, for 100 claims in total.

A bullet with several related numbers counted as one entry because those numbers shared one citation path. Repeated claims were not removed, because the unit was the published claim a reader encounters. Each claim was coded for its cited destination, whether the destination was usable, whether it was an original producer, and whether claim-specific methodology was publicly visible.

A publisher's own survey or platform analysis counted as original. Government data counted as original. A proprietary market-research company counted as the producer of its own estimate. Statista and report marketplaces were coded separately when the public page worked as an aggregator or storefront rather than exposing the originating evidence. A link to a producer counts as original even if the exact figure is no longer visible, provided the producer created the relevant research; where the producer or research could not be established, the entry is coded as no usable source.

What are the limits of the findings?

• The findings describe these 100 claims. They are not a probability estimate for all PR content online.
• Publisher scores cover only the first 20 eligible entries on each sampled page, not the whole page or the publisher's work.
• Live pages and report landing pages can change after capture.
• The audit measured citation pathways and public method visibility. It did not independently reproduce every survey or market estimate.
• Figure confirmation was completed for only part of the original-source entries, and 5W coded the entries without an independent second coder.
• A paywalled original producer received original-source credit, but not methodology credit unless public method details were visible.

The central claim is narrow on purpose: in this 100-claim sample, fewer than half of the statistics linked directly to the organization that produced the underlying research or data. The audit did not test whether the statistics are false.

How should this audit be cited?

5W Research, PR Statistics Citation Audit, October 27, 2026. https://www.5wpr.com/research/pr-statistics-citation-audit/

Frequently asked questions

How many PR statistics linked directly to original research?

Forty-six of the 100 audited claims linked directly to the organization that produced the survey, dataset, platform analysis or market estimate.

Does this audit say PR statistics are false?

No. The audit tested whether readers could trace a number to its producer in one click, not whether each number is accurate.

Which page linked to original sources most often?

Sprout Social linked 19 of its 20 sampled entries directly to original producers. BlueTree linked 17 of 20, PRLab 7, Scopic Studios 2 and flair 1. These scores cover only the first 20 eligible entries on each page.

How many claims linked to a source with some claim-specific methodological detail?

Thirty of the 100 claims. A destination earned credit if it publicly showed any claim-specific sample detail, field dates, data scope or method, so the 30 does not mean every one disclosed both sample size and field dates. All 30 were among the 46 original-source claims.

How many claims had a usable destination?

Eighty-eight claims led to a destination classified as usable under the audit's rules. The other 12 were one entry with no link attached, four whose links were broken, irrelevant or did not show the cited figure, and seven whose vendor page did not establish the claim's research or producer.

Why does a claim count as original if the exact figure is no longer visible?

The audit's test is whether a reader can reach the producer of the research in one click. Producers revise estimates and update pages, so figure confirmation is recorded separately for each entry in Table 6.

How were the 100 claims chosen?

The first 20 quantitative entries in body order were taken from each of five accessible English-language PR-statistics roundup pages. The sample is purposive, not random.

CONCLUSION

A statistic is only as useful as the path back to its source. Link the original report, show the sample and the dates, and keep the population in the sentence. If your team publishes PR research or statistics pages and wants every number to hold up under a journalist's check, collaborate with 5WPR to build citation-ready content.

5W PR Statistics Citation Audit | October 27, 2026