Flexe.io Research is the editorial desk behind everything published on this site. It exists so that a reader can answer one question before trusting a number here: where did this figure come from, and what happens when it turns out to be wrong.
Who writes this
Articles are written by the Flexe.io team rather than by a named individual columnist. The people involved are the ones running the campaigns: the PR desk that maintains the outlet inventory, the team that vets influencer accounts, and the analysts who reconcile campaign results. Attributing that work to one invented byline would misrepresent how it is produced, so we do not.
Where the numbers come from
Performance figures are first-party. They come from our own campaign database: 150 campaigns run between January 2024 and February 2026, measured over a 90-day window using UTM data reconciled against on-chain activity. The median net return across that set is 2.8x, with a range of 1.4x to 7.0x by project type. The 12% failure rate is included in that median rather than excluded from it, which is the difference between a benchmark and a showcase.
Third-party figures are labelled as such and dated. Outlet readership comes from Similarweb. Domain authority and trust metrics come from commercial link databases. Where a figure is an estimate rather than a measurement, we say so in the sentence rather than in a footnote.
What happens when we are wrong
Published numbers decay. Ours have, repeatedly, and the corrections are the part of this page worth reading.
- Outlet authority scores. A re-audit of our PR inventory against a third-party export found 53 outlets whose published score was wrong. Nineteen were off by five points or more, and sixteen of those nineteen had been overstated. All were corrected, and the post-correction check found zero remaining discrepancies across 108 values.
- Forum traffic. We published a monthly readership figure for a forum campaign that was roughly three times the real number. It had been measured once and never rechecked. The page now carries the corrected figure with the date it was measured.
- Blockchain explorer traffic. Advertising inventory figures collected in late 2024 were still on the site in 2026, overstating several explorers by a factor of three to four. Replaced and dated.
The common cause in all three cases was the same: a number gathered once and treated as permanent. The rule that came out of it is that any inventory update re-checks both traffic and authority, not whichever one prompted the update.
What we publish about our own inventory
We publish the parts that are inconvenient. Our study of which crypto PR outlets pass a followed link names the seven outlets in our inventory whose authority score sits above 50 while their trust score sits below 25, including the two that we sell inside the followed-link group. It also states plainly that only 30 of 195 release outlets carry a followed link, which is a weaker number than the industry usually quotes.
The reasoning is straightforward. An agency that only publishes flattering data about its own inventory has told you nothing you can check.
What we do not publish
- Client-identifying results without that client’s permission. Named cases on this site are published with consent.
- Guaranteed outcomes. We publish medians and ranges, not promises, and we list what cannot be bought at all.
- Figures we cannot source. If a claim is circulating in the industry without an origin, it does not appear here, even when it would help the argument.
Method notes and corrections
How net return is calculated, what the 90-day window includes and excludes, and how campaigns enter the benchmark set are documented on our methodology page. If you find a figure on this site that you believe is wrong, tell us and we will check it. Every correction listed above started that way or started with our own re-audit, and both routes are open.