Honest dataBlog1 Jul 2026

from the DAAC journey, July 2026

Public data ≠ usable data (real ≠ symbol)

Information can be public and still be unusable: scattered, unverified, described five different ways by five sources that never spoke. The gap between the two is the actual job.

Ask whether information about a sector is public, and you will often get the wrong answer to the question that actually matters. Yes, it's public: it sits somewhere on a government website, a report PDF, a registry nobody consolidated. And no, in every practical sense, it isn't usable by anyone trying to make a real decision. Scattered across formats, undated, uncorroborated, described in five vocabularies by five sources that never talked to each other. Public and usable are not the same claim, and the gap between them is where we work.

We were named on the first of July for the sector-agnostic engine that exists to close that gap: scan the public record, verify it, organise it, keep it current. Not build a single reference site once and walk away. Maintain the record the way any living thing has to be maintained, because information decays the moment nobody is checking it anymore.

We are careful about the difference this post's title is pointing at, because it is the whole discipline: a symbol is not the real thing. A name appearing in a document is a symbol that a real, verifiable fact might exist behind it. It is not proof the fact does. Confusing the two is the single most common failure mode of anyone claiming to have "the data" on a sector: they have found where the symbol lives, and stopped, without ever confirming what stands behind it. Our job starts exactly where that stopping point usually is. We do not scan a source and take its claims at face value. We corroborate, and we say plainly, on every record, what was actually checked: not "verified" as a blanket badge, but which specific claim was checked against which specific independent source, and what was not.

We only take on a sector under two conditions held together: the information has to be genuinely opaque, meaning it is hard to find and harder to trust once found, and there has to be no existing shared record anyone already maintains that people can rely on instead. Where a good maintained registry already exists, building a second one is waste, not service. Where neither condition holds, whole sectors of real, decision-relevant information sit unusable, not because nobody wrote it down somewhere, but because nobody turned what was written down into something a person could actually act on.

Our first hub went live at the end of June, and what it proved mattered more than the hub itself: a hub is not a static deliverable, it is one output of a repeatable method. Select a sector, scan it, verify it, organise it, keep it maintained, and the same method scans a different sector next. A hub built this way turns out to play more than one role once it exists. It is a public front door anyone can walk through. It is a stream of verified supply the rest of the suite can build on. And, less obviously until you watch it happen, it becomes the ground a sector's own community starts to gather around, because for the first time there is one shared record worth contributing to instead of a dozen private ones nobody else can see.

Public was never the hard part. Anyone can point at a source and call it open. Usable is the hard part, and it is the only part that actually matters to whoever is trying to make a real decision with what they find.