A list cleared verification at a rate any deliverability policy would accept. Then a rep opened the first record and could not answer one question: why this person, this week.

That is not a data-quality problem in the usual sense. It is a qualification problem, and it does not appear in any dashboard built to count bounces.

The evidence position

Two limits belong at the top, because the argument rests on what can actually be checked.

First, vendor-published studies on B2B list decay circulated widely this week. Their pages sit outside this publication's approved source set and did not verify against it, so no decay percentage, half-life or invalid-rate figure from any of them appears below.

Second, the four documents cited here were not returned to this edition's automated retrieval. They are cited only for narrow, stable propositions about what each one is and what it requires. No statistic is quoted from any of them.

What follows is analysis, and it is labeled as such where it matters.

Two records that score identically

Record A: verified address, current employer, a title that has not been right for nine months, and no event of any kind attached to the account.

Record B: verified address, current employer, a title confirmed this quarter, and a filing dated last Tuesday that changes what the company is likely to buy.

A validation dashboard scores these the same way. Both addresses will accept mail. Both clear a suppression list. Neither will bounce.

Only one produces a conversation, and the dashboard has no column for the difference.

Two clocks sharing one label

The reason teams conflate these is that both get called data quality. They are not the same decay process.

Contact validity asks whether the mailbox exists. It changes when someone leaves, when a domain migrates, when a company closes. It announces itself: a hard bounce tells you.

Reason validity asks whether there was ever a justification for writing to this person, and whether it is still true. It changes with no announcement at all. Nothing bounces when a company stops needing what you sell.

The second failure is silent, which is why it survives so many cleanups.

What turnover data can and cannot tell you

The U.S. Bureau of Labor Statistics publishes the Job Openings and Labor Turnover Survey, which reports hires, separations, job openings and quits at the establishment level. It also publishes an employee tenure series reporting median years of tenure with the current employer.

Both are useful for exactly one thing: establishing that contact records decay as a structural fact rather than bad luck.

Neither can tell you which record just decayed. A separation rate describes a population of establishments. Your list is a set of specific people, and the question a rep needs answered is which one of them changed, on what date.

That is the honest boundary of the statistic. It justifies building a maintenance process. It cannot prioritize one record over another.

The one field most systems do not have

Open a record and you will find a last-modified date. That is a fact about your database, not a fact about the prospect.

What is usually missing is the reason: why we believed this person had a reason to act, where we saw it, and when we saw it.

Without that field, a pipeline review cannot distinguish a live opportunity from an old hope. The reviewer asks a person, the person says something about the spring, and the record survives on memory.

Why a filing is different from a hunch

A filing is a dated fact with an address. The SEC operates full-text search across EDGAR filings, which means a claim about what a company disclosed can be checked by someone who was not in the room.

That property gives the signal three useful traits: dated, sourced and checkable. A signal that cannot be dated cannot be aged, and a signal that cannot be aged cannot be retired.

Filings are one class of dated evidence. Hiring activity is another. Neither is a promise that anyone will buy. Both are things a reasonable person can verify before spending an hour on a call.

The accuracy obligation is not a preference

There is a legal dimension most outreach discussions skip. Article 5(1)(d) of the General Data Protection Regulation requires that personal data be accurate and, where necessary, kept up to date, with reasonable steps to erase or rectify inaccurate data without delay.

Read precisely, that is an obligation about the record, not about the campaign. A stale job title is inaccurate personal data held for a purpose that depends on the title being right.

This is EU law and applies to processing within scope, not to every list everywhere. The portable point is narrower: accuracy is a maintenance duty, and a quarterly cleanup does not discharge a continuous one.

Four questions per record

Before a record enters a sequence, someone should answer these without asking a colleague.

  • Why this person? Not the department — the role, and the authority that role actually carries over the decision.
  • Why this company, now? A dated event, not a fit score.
  • Where is the evidence, and what is its date? A source and a date, or the honest answer is nowhere.
  • What would make this wrong, and would we notice? If nothing would tell you, the record has no maintenance path.

Three delivery options

Buy more volume. Cheapest per record and most expensive per conversation. It fixes a shortage of contacts and does nothing about a shortage of reasons.

Enrich what you already have. Buys accuracy — better titles, better firmographics — and improves deliverability. It still does not create a reason to call this week.

Gate outreach on a dated reason. One required field, enforced where records enter a sequence, carrying a source and an evidence date. Slowest to adopt, and the only one of the three that changes what a rep says on the call.

Most teams need the second to keep the list usable and the third to make it worth using. The first is a decision about volume, not quality.

A 30-day pilot

  • Days 1 to 5. Pick one segment and one offer. Not the product line.
  • Days 6 to 12. Take 50 records from the active queue and try to produce a dated reason for each, from records alone.
  • Days 13 to 18. Score the gaps. Count how many had a verified address and no reason at all. That number is usually the finding.
  • Days 19 to 25. Add one required field — evidence and evidence date — enforced at the point a record enters a sequence, for that segment only.
  • Days 26 to 30. Have someone who did not build the process reconstruct why three records were contacted. If it requires an interview, the record is not finished.

Four indicators worth baselining

Share of active records carrying a dated reason. Age of the newest evidence per record, in days. Attempts per record, since a list with no reasons generates repetition. Reconstructions that required asking a person, which is the honest test of whether the record works.

Baseline all four first. None requires a vendor benchmark, and all four describe your operation rather than someone else's.

What remains unestablished

  • Any measured decay rate for B2B contact data. None is asserted here.
  • Whether signal-based targeting improves reply or meeting rates, and by how much. This edition cannot test that.
  • Which classes of filing or hiring event reliably precede a purchase decision.
  • Whether the same pattern holds outside the United States and under different data-protection regimes.
  • Market data. None was reviewed for this edition, so no stock movements are reported.

Evidence notes and limits

The Bureau of Labor Statistics sources are cited for narrow descriptions only. The Job Openings and Labor Turnover Survey reports hires, separations, job openings and quits at the establishment level, and the employee tenure release reports median years of tenure with the current employer. No rate or figure is quoted from either, because neither page returned content to this edition's automated retrieval.

The SEC's EDGAR full-text search is cited as a search interface over filing text, which is the basis for the claim that a disclosure can be checked by a third party. No filing, issuer or disclosure is characterized here.

Article 5(1)(d) of the GDPR is cited for its accuracy principle as described in the text: personal data accurate and, where necessary, kept up to date, with reasonable steps to erase or rectify inaccurate data without delay. It is EU law applying to processing within scope, and the article says so.

Vendor-published research on B2B list decay circulated this week and informed the framing only. Those pages sit outside this publication's approved source set and did not verify against it, so no percentage, half-life or invalid-rate figure from them appears anywhere in this article.

The two records, the four questions, the three options, the pilot and the indicators are this publication's analysis and suggested method. They have not been validated as a protocol, and nothing here establishes that any organization improved reply rates, reduced waste or increased revenue.

The advertisement above is separate from this analysis and is not a source for any claim in it.