Data decay definition

Search for a B2B data decay rate and the same few numbers appear on page after page, usually without a source. This page traces where the most copied one comes from, replaces it with rates from official U.S. labour statistics, shows which fields decay fastest, and covers how to counter decay without buying the same data again.

What is data decay?

Data decay is what happens to a list after you stop touching it. Every contact record describes a person at a moment in time: their employer, their title, their work email, their direct line. When the person moves, the record keeps describing the old moment. Nothing in the database flags it; the record simply becomes wrong.

In B2B data the main cause is people changing jobs or roles. Smaller causes are company events such as rebrands, mergers and domain changes, and mailbox policies such as accounts being closed when an employee leaves. Decay is not the same as bad data entry. A record can be entered perfectly and still decay, which is why decay has to be measured as a rate over time.

How fast do B2B contacts actually change?

Two official U.S. series give the speed. The BLS Employee Tenure in 2026 release puts median tenure with the current employer at 4.1 years, with large differences by age. The monthly JOLTS release reports how many workers leave a job each month: total separations ran at 3.2 % per month and quits at 1.9 % per month in August 2026.

Both series measure people leaving employers. They do not measure whether a contact record or email address still works, and they do not count role changes inside the same company.

Read the two monthly rates as a range. Quits are people choosing to leave; total separations add layoffs and other exits. A list skewed towards younger staff will move faster than the median, and a list of senior executives slower. Role changes inside a company come on top, and they break targeting even when the email still works.

Two primary sources cover the speed of job change. The BLS Employee Tenure release, published every two years, reports median years with the current employer by age, sex and industry. The monthly Job Openings and Labor Turnover Survey reports hires, quits and total separations as a share of employment. Neither measures data accuracy directly: both describe people moving, which is the main input to decay. Vendor decay rates should be checked for a named study, sample and date before use.

Where does the 2.1 % per month figure come from?

The most repeated decay figure is 2.1 % per month, often annualised as 22.5 % per year. It traces back to HubSpot's Database Decay Simulation, which attributes it to Marketing Sherpa without a link to a study, a sample or a date. None was found. The figure has been copied so often that it now reads as a standard, but it has no primary source that can be checked.

Two other figures travel the same way: about 30 % a year, attributed to Dun & Bradstreet, and about 70 % a year, attributed to ZoomInfo. Neither appears on the publishers' own sites. Treat all three as red flags: if a business case rests on one, replace it with the BLS rates above or with a measurement on your own list.

Which fields decay fastest?

Fields tied to the person decay first, and they decay together. When someone changes employer, their company, title, work email and usually their direct phone change on the same day. Fields tied to the company move more slowly.

• Employer and job title: fastest. They change with every move and every promotion.
• Work email: usually stops working soon after the person leaves, which is when it starts producing hard bounces.
• Direct phone: mobile numbers often follow the person; desk numbers stay with the company and reach a stranger.
• Company attributes: headcount, technologies and open roles drift with hiring and projects, over months.
• Company name and domain: slowest. They change with rebrands and acquisitions.

The pressure is not easing. LinkedIn's Work Change Report (2025) projects that people entering the workforce now will hold about twice as many jobs over their careers as earlier generations.

How do you counter decay without re-buying data?

You cannot stop decay, but you can stop acting on decayed records. The cheapest fixes work at the moment of use, not on the whole database.

1. Pull from a source that refreshes. A database refreshed monthly starts every list closer to the present than a file bought once.
2. Verify before you send. Run email verification on a list right before a sequence starts, not when the list was built.
3. Re-enrich what you will use. Refresh title and employer for the segments in active campaigns or open deals, not the whole CRM.
4. Track job changes as signals. A known contact leaving a customer is an event to act on, not a record to fix at renewal.
5. Close the loop on bounces. Every hard bounce marks a decayed record; update or retire it the same day.

Frequently asked questions

What is a realistic B2B data decay rate?

No checked study measures contact-record decay directly. The best proxy is U.S. labour data: 1.9 % of workers quit and 3.2 % leave a job in total each month (BLS JOLTS, August 2026). Your list will move faster or slower depending on seniority and age.

Is the 2.1 % per month figure reliable?

It has no checkable primary source. It traces to a HubSpot simulation citing Marketing Sherpa without a study, sample or date. Use official labour statistics or measure your own list instead.

Which contact fields go out of date fastest?

Employer, job title and work email, which all change on the day someone moves jobs. Company-level fields such as size or technologies drift more slowly, and company name and domain are the most stable.