Measure Lead Data Freshness by Field Type

A record edited yesterday can still contain a role claim last checked two years ago. Conversely, a stable identifier may remain usable without daily research. A single record-level freshness percentage hides both situations.

Measure freshness at the field and intended-use level. Start with the claims that could change the next decision, then define what a meaningful check would establish. The result should direct verification effort, not encourage people to touch every record so its timestamp looks recent.

Distinguish four dates

Keep the underlying event or effective date, the source publication date where known, the date someone observed the source, and the date the claim was actually reviewed. These can be different. Opening an old announcement today does not make its statement current.

Do not substitute an export timestamp, file modification date, or CRM edit date for verification. If the source date is unknown, retain that uncertainty. A field can have a recent review outcome of unresolved, which is useful but not equivalent to currently supported.

W3C's W3C: Data Quality Vocabulary treats quality in relation to the data's use and provides a vocabulary for quality measurements. The matrix below applies that contextual approach; it is not a universal expiration schedule prescribed by W3C.

Build a field-level review matrix

For each field, record the claim, its consequence, acceptable evidence, review trigger, and resulting state. Use states such as supported for this use, verification due, unresolved, contradicted, and not required. Avoid turning all of them into a single average score.

A fictional account handoff uses these rules:

• Legal entity identifier: check issuer and entity match at creation, then recheck when identity evidence conflicts or a relevant structural change is reported.

• Current professional role: review before relying on it in the planned handoff or communication, and when a departure or role-change signal appears.

• Operating location: verify for the specific service-coverage decision; a planned opening remains separate from an operating site.

• Project milestone: use the latest relevant project evidence and its effective date, not a generic company-page check.

• Contact restriction: consult the authoritative handling record at the action gate; an old restriction does not expire merely because its date is old.

The final row is deliberately different. Freshness logic must not turn a durable stop instruction into permission to contact again.

Calculate a transparent fictional report

Suppose the upcoming handoff requires role evidence for ten records. Six have supporting reviews under the team's stated rule, two are due for verification, one is unresolved after a current check, and one has no usable review date.

The report says six of ten required role fields meet the review rule. It separately shows two due, one unresolved, and one date unknown. It does not say nine of ten are fresh because nine contain a date, or seven of ten are current because the unresolved case was checked today.

For operating locations, only eight of those accounts require the field. Five are supported, two remain planned rather than operating, and one is unresolved. Use eight as the relevant denominator, not ten. Preserve the two planned locations as meaningful findings rather than blank cells.

These invented counts demonstrate reporting logic only. They are not a recommended quality threshold or evidence that any real dataset has a particular decay rate.

Handle a change that affects only one claim

Imagine the company announces that one facility will close next month. The researcher records the announcement and planned effective date, then flags the service-coverage claim for the relevant decision. The company's identifier and unrelated sites are not automatically invalidated.

If the closure later becomes effective, update the operating status with evidence. Do not replace the original announcement date with the day the team noticed it. Retaining the sequence helps explain why earlier research supported a different conclusion.

Use the report to prioritize, not to promise certainty

Sort verification work by the next use and consequence of error. A role claim scheduled for a message needs attention before an optional industry label in dormant planning data. A contradictory identifier can block the entire handoff even when every other field is recent.

Review the rules when you learn which claims change or fail in practice. Keep rule versions so a stricter standard does not look like sudden data deterioration. The useful output is a field-specific queue and an honest readiness statement, not a blanket promise that the whole database is fresh.

Continue with Measure Sales Navigator Prospecting with a Small, Honest Funnel or Compare Two Target Segments with a Fair Sales Experiment.