B2B Data Decay Study: What the Published Data Measures

A B2B data decay study has to follow the same contact records over time and count how many stopped being true. We reviewed every decay and job-change source we could open, and none does that for every field over a full year.

Official labor data, business closure data, email verification data and two vendor studies each measure one part of it.

Below: what each source actually measured, why the popular “30% a year” and “2% a month” figures are hard to trace, and the method for the study we are running on our own data. For what decay is and how to fix it, read our guide to B2B data decay.

What does the evidence on B2B data decay actually measure?

The published evidence on B2B data decay measures proxies: how long people stay with an employer, how often jobs end, how often business locations close, and how many submitted email addresses fail a check. Each is real and sourced. None is the share of your contact records that went stale in a year.

Two vendor studies do measure B2B contacts. Lusha counts job changes recorded in its own database, and Cleanlist re-checked 5,000 CRM contacts every week for 13 weeks. Both publish their method, and both stop short of a fixed 12-month group checked on every field.

How we reviewed the evidence

We read every source on the publisher’s own page, PDF or data file on 27 and 28 September 2026, and followed each cited figure back to the earliest document we could open. Last reviewed: 28 September 2026.

Scope. Any published number on job changes, job tenure, business closures or contact data going bad that sales and marketing pages cite as a decay rate.
Three questions per source. Does it follow the same people or records over time? Does it count B2B contacts? Are the method and sample published?
Primary data only. Where a blog cited another blog, we followed the chain. Figures we could not trace are named in their own section, never used as findings.
Definitions. A stale record has at least one field that stopped being true: employer, title, email or phone. Tenure is time with the current employer on the survey date. A separation is one job ending, counted by the employer.
Limits. The official statistics are US-only. Vendor studies describe their own databases or customers. Gated reports, such as Validity’s full 2025 CRM report, were not read.

Finding 1: no source measures all of B2B data decay

No source we reviewed follows a fixed group of B2B contacts for a year and checks every field. The scorecard shows what each one does, and where it stops.

The official sources are rigorous but count workers and businesses; the two that count B2B contacts are vendor studies with narrow scopes.

The official sources are large and rigorous, but they count workers and businesses, not contact records. The two sources that count B2B contacts are vendor studies built on one company’s database or one customer’s CRM. The two most quoted rates, “30% a year” and “2.1% a month”, have no study behind them that we could open.

Finding 2: vendor studies put job changes near 1% a month for US sales leaders

Lusha, a B2B contact data vendor, found that 12.25% of 140,284 US sales leaders at VP and C-suite level had a recorded job change over 12 months, and 25.67% over 24 months. Both work out to about 1% a month (Lusha, run of September 12, 2026; the first run in August found 12.6%).

Region mattered most. UK sales leaders changed roles at 21.40% over 12 months, 1.75 times the US rate. Seniority barely moved the number.
A floor, not a ceiling. Lusha counts only changes its database has detected, and it measured today’s group looking back, not a group fixed a year earlier. It says both on the page.
Detection lags. Over 3 months the rate was 0.50% a month, half the longer-run pace, which Lusha reads as changes taking time to surface.

Cleanlist, a data enrichment vendor, took 5,000 contacts from an anonymized customer CRM (North America, several industries), re-ran each record through its enrichment tools every Monday for 13 weeks, and counted records where title, company, email validity or phone connectivity changed. It reported 1.8% to 2.4% a week and compounded that to about 67% a year (Cleanlist, Q1 2026).

Read the two side by side with care. Lusha counts one event, a job change, in one segment. Cleanlist counts any change in its tools’ output, which can include a different answer for a person who did not move, and projects 13 weeks to a full year.

Finding 3: about 1 in 5 US workers is in their first year with an employer

20.6% of US wage and salary workers had been with their current employer 12 months or less in January 2026, down from 22.2% in January 2024 (BLS Employee Tenure in 2026). Short tenure falls steeply with age.

Senior buyers move far less often than junior staff, so the age mix of a list sets how fast it ages.

Mid-career workers move less. 15.4% of workers aged 35 to 44 and 12.2% of those aged 45 to 54 had a year or less with their employer, against 25.5% at 25 to 34.
Not a pure job-change rate. BLS says the short-tenure group includes new hires, people who lost a job and found another, and voluntary job changers. First jobs are in it too.
Median tenure was 4.1 years in January 2026, up from 3.9 years in 2024.

Lifetime data points the same way. Americans born 1957 to 1964 held an average of 12.9 jobs between ages 18 and 58 (BLS National Longitudinal Survey of Youth 1979, 9,964 people followed since 1979).

Even among jobs started at ages 45 to 54, 21% ended in less than a year. A job that just started is the least stable one on your list.

Finding 4: job churn peaked in 2022 and has cooled since

US employers recorded 62.8 million separations in 2025, down from 72.3 million in 2022. Quits fell from 50.5 million to 38.0 million over the same years (BLS Job Openings and Labor Turnover Survey, 2025 annual estimates).

Job churn peaked in 2022; in 2025 US employers logged about 10 million fewer separations than at the peak.

The average monthly separations rate eased from 3.9% in 2021 and 2022 to 3.3% in 2024 and 2025. Separations are events, not people: one person can leave several jobs in a year, so these counts cannot be read as the share of your contacts who moved.

The rate varies by industry, which suggests a list of agency or hospitality contacts ages faster than a list of bankers or manufacturers.

Industry	Average monthly separations rate, 2025
Leisure and hospitality	5.6%
Professional and business services	4.6%
Construction	4.0%
Retail trade	3.8%
Health care and social assistance	2.9%
Information	2.8%
Manufacturing	2.4%
Financial activities	2.3%
Government	1.5%
All US nonfarm employers	3.3%

Source: BLS JOLTS, Table 20, annual average of monthly rates, not seasonally adjusted. BLS revises annual estimates each January.

Finding 5: about 9 in 100 business locations close each year

9.4% of US employer business locations closed in 2023: 699,980 of them, against 790,295 that opened (Census Bureau Business Dynamics Statistics). Over the decade before, the exit rate stayed between 8.4% and 10.1%.

About one business location in eleven closes every year, and the rate has barely moved in a decade.

Firm closures jumped. 523,958 employer firms closed all their locations in 2023, up from 450,039 in 2022.
Mergers are missing. Census does not count a firm that disappears in a merger or acquisition as a closure, so company churn in a B2B database, from rebrands, acquisitions and domain changes, runs higher.
Employer businesses only. Firms with no employees are not in these counts.

Finding 6: email data measures lists sent for cleaning, not lists aging

ZeroBounce found at least 23% of the 11 billion+ email addresses it verified in 2025 were invalid or risky, down from 28% in 2024. Only 62% of submitted addresses were valid (ZeroBounce Email List Decay Report, 2026).

Each point is a different set of lists sent for cleaning, so the line shows how dirty submitted lists are, not how fast one list ages.

Each year’s figure describes different lists: whatever customers chose to submit, B2B and consumer mixed. That is how the line can fall from 28% to 23% in one year while job churn barely moved. It is a fair warning about unverified lists, not a decay curve for one list.

Two more data points fill in the picture:

Bounces. Across 31 million cold emails sent through Hunter in 2025, the average bounce rate was 3.6%, which Hunter traces mainly to unverified contact data.
Perception. 48% of 631 CRM admins told Validity in 2024 that customer data was decaying faster than a year earlier. That is what admins noticed, not a measured rate.

Bounce benchmarks and sender rules are in our email verification statistics.

Why the “30% a year” and “2% a month” figures are hard to trace

The two most quoted B2B decay rates have no original study we could open. Here is where each trail ends.

Figure you will see	Where the trail ends	What is missing
“B2B data decays 30% a year”	No original found, by us or by Lusha, who also looked. One vendor report credits a 2017 Harvard Business Review article, but that article measured errors in 75 managers’ data samples and gives no yearly decay rate.	Any study, sample or method
“2.1% a month”, or 22.5% a year	HubSpot’s Database Decay Simulation, citing MarketingSherpa research. The MarketingSherpa study could not be found, and HubSpot’s page is undated, about a decade old.	The original study
“70% a year”	A 2011 MarketingSherpa article quoting a Salesforce executive: “as high as 70 percent”.	A measurement: it is an upper-bound opinion

The HBR article does contain a 30%, but it describes quality scores: a quarter of the data quality scores in its sample fell below 30%. It says nothing about how fast records go stale.

Two details make these numbers easy to mix up. Monthly rates compound, so 2.1% a month is 22.5% a year, not 25.2%. And Lusha argues the popular 30% looks like a two-year job-change rate: its own 24-month figure for US sales leaders is 25.67%.

The full list of untraceable decay numbers, including “70.3% per Gartner”, is in the guide linked at the top of this page.

Our B2B data decay study: the method

The decay study we are running will follow one fixed group of B2B contacts for 12 months and re-check every field at 3, 6, 9 and 12 months. We publish the method before the results, so it cannot be adjusted to fit them.

Following the same records forward is the one design none of the published sources uses for every field.

Fix the cohort on day 0. A random sample of US contacts from Lead Finder searches, each returned with a Valid or Risky work email, a mobile number, or both. Planned size: 2,000 contacts, 500 in each of four job-function groups (sales, marketing, operations, IT and engineering). At 500 per group, a 20% stale rate carries a margin of about 3.5 points either way at 95% confidence (our arithmetic).
Record the baseline. Employer, title, email status, mobile number and the date of each check.
Re-check at 3, 6, 9 and 12 months. Re-verify the email. Re-find the person with a fresh lookup and confirm the employer and title. Re-run the mobile lookup and record whether the same number comes back for the same person.
Count what changed. A record is stale when the employer changed, the title changed at the same employer, the email no longer verifies, or the mobile changed. We report each field on its own and “any field”, with confidence intervals, and we never compound a short window into a yearly rate.
Publish. Aggregate tables, this method, and an anonymized file with no names, emails or phone numbers.

What the design cannot do, stated up front:

It starts from Scalelist results, so it describes contacts Scalelist can find, not every B2B contact.
Risky stays uncertain. A catch-all domain accepts any address, so an email that stays Risky tells us nothing new. We report those separately.
A returned mobile is not proof it rings the person. Only a call proves that, and we will not call people just to test a number.
US first. Lusha found region was the strongest driver, so we will add a region only when it can reach the same sample size.

One question no source we reviewed answers is whether mobile numbers outlast work emails after a job change. For every person who changed employer, this design records whether the email and the mobile survived.

What this means for your team

Until a fixed-cohort study exists, plan with ranges by segment and measure your own list.

Rank lists by how fast they are likely to age. Younger contacts, and industries such as professional and business services or leisure and hospitality, turn over faster than senior staff in finance or manufacturing.
Treat a recent job start as a risk and a signal. New jobs are the least stable ones, and a buyer who just moved is also a reason to reach out: see job changes as a pipeline source.
Verify at the moment of use. Databases detect job changes late, so a recent “last updated” date is weaker evidence than it looks. Checking before a sequence is one of the prospecting techniques that still work.
Measure your own decay. Re-check a sample of records verified on a known date; the guide linked at the top walks through a 200-record method. When you compare vendors, test them on your own accounts.
Fix decay where it costs most. Every bounce and wrong-person call traces back to the list, the first step of sales prospecting.

Re-check a sample of your oldest contacts and see which emails still come back Valid.

Get 20 credits to test it

How to cite this page

Cite this page as: Scalelist, “B2B data decay study: what the published data measures”, scalelist.com/research/data-decay/, last reviewed 28 September 2026. When you quote a single number, name its original publisher (BLS, Census, ZeroBounce, Lusha or Cleanlist) and its year.

FAQ

Is there a study that measures B2B data decay?

Not a complete one. No B2B data decay study we reviewed follows a fixed group of contacts for 12 months and checks employer, title, email and phone. The closest are two vendor studies: Lusha counted recorded job changes for 12.25% of US sales VPs and C-suite leaders over 12 months, and Cleanlist re-checked 5,000 CRM contacts weekly for 13 weeks.

What is the B2B contact data decay rate per year?

There is no single verified yearly rate. The measured pieces: 12.25% of US sales leaders changed jobs over 12 months in Lusha’s database (2026), 20.6% of US workers had been with their employer a year or less (BLS, January 2026), and 9.4% of US business locations closed in 2023 (Census). Your rate depends on roles, industries and regions.

Where does the “30% of B2B data decays every year” figure come from?

We could not find an original study, and Lusha reported the same in September 2026. One vendor credits a 2017 Harvard Business Review article, but that article measured errors in 75 managers’ data samples and gives no yearly decay rate. Lusha’s own count suggests 30% is closer to a two-year job-change rate for US sales leaders.

Is 2.1% a month the same as 25% a year?

No. Monthly rates compound, so 2.1% a month is about 22.5% a year, not 25.2%. The 2.1% figure comes from HubSpot’s Database Decay Simulation, which cites MarketingSherpa research we could not find. It is about a decade old and has no published method, so treat it as a rule of thumb, not a measurement.

How many people change jobs each year in the US?

No official statistic counts it directly. The closest: 20.6% of US wage and salary workers had been with their employer 12 months or less in January 2026 (BLS), a group that also includes first jobs. Employers recorded 62.8 million separations in 2025 (BLS JOLTS), but those are events, and one person can have several.

Which B2B contacts go stale fastest?

Younger workers and high-turnover industries. 25.5% of US workers aged 25 to 34 had a year or less with their employer in January 2026, against 12.2% at 45 to 54 (BLS). Professional and business services averaged a 4.6% monthly separations rate in 2025, against 2.3% in financial activities. In Lusha’s data, UK sales leaders moved 1.75 times as often as US ones.

When will Scalelist publish its own decay results?

After the first re-check of the fixed cohort, with more results at 6, 9 and 12 months. The method is on this page now, and each release will add the sample, dates and an anonymized file with no personal data. Until then this page contains no Scalelist decay rate, and every number on it comes from the publisher named next to it.

Overview

use cases

Features

Resources

B2B data decay study: what the published data measures, and how we will measure it