Evaluating Confidence in Property Data You Did Not Collect
You will spend most of your career acting on data other people gathered. This is a working framework for judging it: weigh recency, source authority, and corroboration, and have a rule ready for when records conflict. The goal is not certainty — it is a defensible reason for treating one value as stronger than another.
How do you evaluate confidence in property data you did not collect?
Judge each value on four axes: how recent it is, how authoritative its source is, whether an independent record corroborates it, and — when records disagree — which one your conflict rule favors. A value that is current, drawn from a strong source, and matched by a second record deserves high confidence. A value that is old, singular, or from a weak source deserves caution regardless of how convenient it is.
Inherited data arrives without the context its collector had. You did not see the register it came from or the day it was pulled, so you cannot lean on memory or intuition. That absence is precisely why an explicit framework beats gut feel: it forces you to interrogate every value the same way, and it gives you language to explain to a committee why you trusted one figure and discounted another.
The four axes of confidence
- Recency — how close the value's observation date sits to today, since property facts drift as owners change and structures are altered.
- Source authority — whether the value traces to an official public record, a documented provider, or an unverifiable secondary mention.
- Corroboration — whether an independent source reports the same value, which raises confidence more than any single record can on its own.
- Conflict resolution — a predetermined rule for which source governs when two credible records disagree, applied before emotion or convenience intervenes.
These axes are not equally weighted in every situation. For a fast-moving field like ownership, recency dominates; a slightly less authoritative but current record can beat a pristine but stale one. For a structural fact that rarely changes, authority and corroboration matter more than the exact capture date. Naming which axis governs a given field is part of the judgment, and it should be written down so a reviewer can challenge it.
Resolving conflicts without guessing
A conflict is information, not noise. When two credible sources disagree, resist the urge to average them or silently pick the convenient one. Instead, apply a rule you set in advance: prefer the more authoritative tier; if tiers match, prefer the more recent; if both match, hold the field as contested and route it for verification. The point of a fixed rule is that it removes the temptation to bend a number toward the answer you were hoping for.
| Situation | Confidence signal | Suggested action |
|---|---|---|
| Recent, authoritative, corroborated | Strong | Rely on it, note the lineage |
| Recent but single weak source | Mixed | Seek a second source before relying |
| Authoritative but stale | Mixed | Refresh before a decision depends on it |
| Two credible sources disagree | Contested | Apply conflict rule, flag for review |
Turning the framework into a repeatable check
For each material field, record its source and observation date, ask whether an independent record agrees, and grade confidence from those inputs rather than from how much you like the value. When sources conflict, apply the tier-then-recency-then-hold rule consistently. Any field that cannot clear a minimum bar stays marked unknown instead of being promoted to a fact.
- —It does not turn a weak dataset into a strong one; it only helps you see the weakness clearly.
- —It cannot resolve a conflict where both sources are wrong — some fields simply require primary verification.
- —It is a judgment aid, not a substitute for legal, tax, or investment advice on a specific deal.
Frequently asked questions
Should I ever average two conflicting values?
Rarely, and never for identity or legal-status fields. Averaging invents a third value that neither source reported. Prefer a conflict rule that selects one governing record and flags the disagreement, so the provenance stays honest.
How recent is recent enough?
It depends on the field's volatility. Ownership and occupancy change often and demand fresher observation; stable physical characteristics tolerate older dates. Set the threshold per field type and make it explicit rather than applying one blanket cutoff.
Grade third-party data with a consistent rule.