Premisely is designed so that a score can be challenged: where the evidence came from, when it was checked, how certain it is and which part of it a person decided. This page states the method and, just as importantly, its limits.
The order matters more than any single step. Scoring before the brief exists produces confident nonsense; paying a provider before identity is resolved pays for the wrong record.
Brand DNA turns positioning, audience, occasion and explicit exclusions into named, weighted signals. Weights are visible and editable, and every score decomposes back into them—so a result can always be argued with rather than merely accepted.
A territory is a real geography—region, country, city—not a search radius. Coverage is stated per territory so an empty neighbourhood reads as a gap in the data rather than an absence of opportunity.
Open mapping data, municipal and tourism datasets, official websites and published event programmes create attributable observations before any paid provider is called. Discovery produces candidates, never verified Accounts.
Whether a place exists, whether it is still trading, and who operates it are three different claims with three different evidence bars. Ownership stays Unresolved until direct or corroborated evidence supports it—never independent simply because no group was found.
Brand fit answers whether the Account suits the brand, commercial potential answers whether it is worth the visit, and confidence answers how much the evidence supports either. Fit bands are fixed: 70 and above is strong fit, 50–69 is worth a look, below 50 is not now.
Every claim carries a method—captured, inferred or declared—and a checked date. Inferences below their confidence floor are not shown at all, conflicting Sources open a decision a person records, and qualification and export always need a deliberate human action. Outreach automation is post-MVP.
A methodology that only lists strengths is marketing. These are the things Premisely deliberately will not do, and they constrain the product as much as anything above.
Contacts exist to open a conversation about an Account you already scored. There is no exportable people database, and guessed email patterns stay candidates rather than being marked verified.
Restricted member pages, login-walled content and terms-prohibited endpoints are out of scope. Where a Source needs written commercial permission, it stays a private human lead until that permission exists.
Audience and vibe describe observable occasion, programming and catchment context. Protected demographic traits are never inferred, stored or scored.
An inferred claim is labelled as inferred and capped by the evidence beneath it. A polished number that hides weak evidence is a worse answer than an honest gap.
Every paid provider run states a worst-case estimate in euros and reserves it against the Workspace budget before the first call. Demo World makes no paid calls at all.
Roadmap capabilities are marked as planned wherever they appear. Certifications, regulatory approvals and integrations are not claimed until they are independently established.
This page describes how the pilot works today. It does not claim certifications, regulatory approvals or guarantees that have not been independently established.
Pilot termsDemo World shows how a team explores Account fit without hiding uncertainty or triggering a single paid action.