You are putting together a vendor shortlist, and Adobe is on it. You open the APIs.io estate page for Adobe because it looks like the quickest summary. It says the estate rolls up 7 members and rates 22.1, emerging. The composite average is 26.0, the best member scores 62.7, and the worst scores 3.8. Read on its own, that looks like a thin, scattered API program, and you are about to write that in your notes.
Before you do, the post behind that page says to stop. The catalog's own records tell a different story. Adobe's provider record scores 76.2, exemplar, with 65.9 on Agent Readiness. Adobe Suite scores 79.0, exemplar and agent-native, across 70 published APIs. Adobe Analytics scores 67.6, exemplar and agent-native. None of those three is a member of the estate.
So you look at who is in it. Frame.io is a product at 62.7, strong. Adobe Premiere Pro is a product at 58.7, strong. Rephrase.ai came in by acquisition and scores 27.0, thin. Behance, a product, scores 9.8, minimal. TubeMogul, also a product, scores 3.8, minimal. Frame and Omniture, both acquisitions, are unrated. That last detail is odd, because the overview says all 7 members carry a rating, and two of them don't. Behance and TubeMogul enter as "parent-company-property", with no APIs counted. The estate header counts just 3 APIs in total.
Then you check what the estate leaves out. The catalog holds 15 provider records whose slug starts with adobe, and the estate includes only one of them, Adobe Premiere Pro. The other fourteen sit outside it. They include Adobe Acrobat Sign at 33.2 and Adobe Commerce at 18.6. They also include Adobe Creative Cloud, Experience Cloud, Experience Manager, Photoshop, Launch and Campaign. The parent record itself is outside the estate too.
The post's reading is plain. The estate measures a sample built from acquisitions named in prose and a few declared products. It does not measure Adobe's API program. It keeps the long tail, like TubeMogul with no APIs counted, and drops the records where Adobe actually publishes. The result is a 54-point gap between the estate rating and the parent's own score. The numbers also drift between pages. Frame.io's provider page now reads 65.9, while the estate still shows 62.7.
The turn is that the post does not blame Adobe. It files the problem against the catalog itself. In its words, this is a membership problem in the rollup, not a finding about Adobe. Estate membership should root at the parent and pull in the products it owns, and the adobe-* records are the obvious first candidates. That fix hasn't landed. Until it does, the post says not to quote the 22.1 as Adobe's rating.
That changes what you write in your notes. You don't record Adobe as emerging or thin. You note that the estate figure is a known artifact of how the rollup was built, and you point your evaluation at the records that reflect where Adobe publishes: the Adobe provider page, and Adobe Suite.
The post does name one thing the estate is good for today. It shows which acquisitions never became developer surfaces. If that question matters to your shortlist, such as whether a purchased product ever turned into an API you could build on, the estate answers it. For judging the size and quality of Adobe's API program, it doesn't.
What you take away is simple. A rollup score is only as good as its membership list, so check who is in the estate before you trust the number. Here the catalog told you itself that its headline figure was the wrong one to use. Start from the Adobe provider page and Adobe Suite, and treat 22.1 as a bug report, not a rating.
Read the original: https://apis.io/2026/11/11/adobes-estate-rates-22-1-and-leaves-out-every-adobe-record-that-scores-exemplar/