Verification & brand safety

Ten companies, 159 claims, 46 of them outcome claims: 19 implied, 23 stated metric, 4 guaranteed, 0 none.

What outcome means here is downstream advertiser performance attributed to placements the page says were screened or scored, counted beside the spend kept out of invalid or unsuitable inventory.

Evidence on those 46 claims is 15 bare numbers, 11 named clients, 2 third-party references, 2 described methods and 16 with none; the figures sit on case-study cards and stat tiles, which establishes that a result was recorded, not what it was measured against.

One figure is attributed to a named examiner: Channel Factory's product page reports PwC procedures finding 23% lower media wastage and up to 12% cost savings, with no scope, period or comparison base on the page channelfactory.com.

All four guarantee-coded claims cover cost per view, completion rate, cost savings and attention rather than sales, and no captured page in the segment uses make-good, refund or credit language.

Of 36 pillars, 31 carry evidence somewhere on the site, 16 of those on a page other than the front door, and 3 carry none.

Nothing here travels to a second buyer without a stated base, window and advertiser category; two of the ten sites reach a third party at all, so transfer rests on the vendor's own record.

Open ground: 0 of 10 sites describe how a placement bought by an agent or carried in an AI answer is checked, and the only mentions are two nav categories on one site; boundary is up to six front-door pages per company in one capture week, with DoubleVerify's and IAS's case-study paths 404 and Lumen's home returning animated placeholders.