B2B Data Decay 2026 Statistics (Updated September)

How fast does B2B contact data decay

1. 1

   HubSpot's database decay simulator says email marketing databases lose about 22.5% of their contacts every year.

   Source: HubSpot ^[1]

2. 2

   HubSpot bases that figure on MarketingSherpa research that found B2B data decays at 2.1% a month, an annualised rate of 22.5%.

   Source: HubSpot ^[1]

3. 3

   IVRIS Tech traced the 2.1% monthly figure to MarketingSherpa's 2012 B2B Marketing Benchmark Report, so pages that cite both 2.1% a month and 22.5% a year are quoting one study twice.

   Source: IVRIS Tech, 2026 ^[2]

4. 4

   The 70.3% a year figure that some vendors quote as a decay rate is the share of business cards with at least one changed field after 12 months, from a self-reported survey run in 2002 and updated in 2009, according to IVRIS Tech.

   Source: IVRIS Tech, 2026 ^[2]

5. 5

   Published annual rates differ partly by arithmetic: a 3.6% monthly email decay rate is 35.6% a year when compounded and 43.2% when multiplied by 12, IVRIS Tech shows.

   Source: IVRIS Tech, 2026 ^[2]

6. 6

   RevenueBase, a B2B data vendor, measured business email decay of 3.6% in the single month of November 2024, against a usual rate of 1.5% to 2.0% a month.

   Source: RevenueBase, 2024 ^[3]

7. 7

   Data8, a UK data quality company, puts average data decay at 30% a year: 3,000 outdated records in the first year for a 10,000-contact database.

   Source: Data8 ^[4]

8. 8

   Data8 says 18% of telephone numbers change every year.

   Source: Data8 ^[4]

How fast do email lists decay

1. 9

   0.4% of work email domains in a random sample of 1,000 addresses TAM Graph found since October 2025 no longer have a mail server or website record in DNS.

   Source: TAM Graph · addresses found October 2025 to September 2026, checked 2026-09-24 · 1,000 email addresses · method

2. 10

   ZeroBounce checked more than 11 billion email addresses in 2025 and found at least 23% of them invalid or risky.

   Source: ZeroBounce, 2026 ^[5]

3. 11

   ZeroBounce's yearly email list decay figure was 28% in 2024, 25% in 2023, 22% in 2022 and 23% in 2021.

   Source: ZeroBounce, 2026 ^[5]

4. 12

   More than 9% of the addresses ZeroBounce checked in 2025 were catch-all addresses, which accept mail for any name at the domain and can still bounce.

   Source: ZeroBounce, 2026 ^[5]

5. 13

   Spam traps made up only 0.01% of the addresses ZeroBounce checked in 2025.

   Source: ZeroBounce, 2026 ^[5]

6. 14

   MailCleanup re-checked 912,317 deliverable addresses from one client's list every month for a year and found that 23.79% had decayed.

   Source: MailCleanup, 2026 ^[6]

7. 15

   In MailCleanup's tracked list, decay sped up from 2.79% in the first quarter to 11.91% in the last quarter.

   Source: MailCleanup, 2026 ^[6]

8. 16

   Six months after cleaning, 9.30% of MailCleanup's tracked addresses had decayed.

   Source: MailCleanup, 2026 ^[6]

9. 17

   Deleted mailboxes caused 55.31% of the decay MailCleanup measured.

   Source: MailCleanup, 2026 ^[6]

10. 18

    Deleted mailboxes, disabled mailboxes and expired domains together caused 80.84% of the 217,061 addresses that decayed in MailCleanup's study.

    Source: MailCleanup, 2026 ^[6]

11. 19

    Before MailCleanup's tracking began, 87,683 of about 1,000,000 addresses on the same list already verified as undeliverable, accept-all or unknown.

    Source: MailCleanup, 2026 ^[6]

12. 20

    A study of author contact addresses in MEDLINE, the biomedical literature database, estimated that roughly 2.1% of them become invalid every year, about 69,000 of 3,283,151 addresses.

    Source: bioRxiv, 2019 ^[7]

13. 21

    A 2019 bioRxiv study of 3,283,151 author contact email addresses in MEDLINE estimated that 18% of them were invalid.

    Source: bioRxiv, 2019 ^[7]

14. 22

    32% of all new author email addresses in MEDLINE in 2018 came from free email providers rather than employers or universities.

    Source: bioRxiv, 2019 ^[7]

How often do business contacts change jobs

1. 23

   US wage and salary workers had been with their current employer for a median of 3.9 years in January 2024, down from 4.1 years in January 2022.

   Source: U.S. Bureau of Labor Statistics, 2024 ^[8]

2. 24

   22% of US wage and salary workers had been with their current employer for a year or less in January 2024, down from 24% in January 2022.

   Source: U.S. Bureau of Labor Statistics, 2024 ^[8]

3. 25

   Among US workers aged 55 to 64, 52% had been with their current employer for at least 10 years in January 2024, against 21% of workers aged 35 to 39.

   Source: U.S. Bureau of Labor Statistics, 2024 ^[8]

4. 26

   In an informal survey of 1,025 businesspeople at seminars, IndustrySelect found that 70.8% of their business cards had at least one change in the previous 12 months.

   Source: IndustrySelect ^[9]

5. 27

   In the business-card survey IndustrySelect cites, job titles or functions had changed for 65.8% of people within 12 months.

   Source: IndustrySelect ^[10]

How fast do company domains and web pages disappear

1. 28

   5.1% of company domains marked active in Crunchbase records no longer resolve in DNS, in a random sample of 998 that TAM Graph checked in September 2026.

   Source: TAM Graph · Crunchbase records loaded into TAM Graph August 2024 to September 2026, checked 2026-09-24 · 998 active companies with a domain · method

2. 29

   In a random sample of 998 active-marked Crunchbase company domains that TAM Graph checked in September 2026, 9.6% of companies founded in 2020 or later no longer resolved in DNS (251 checked), against 4.3% of those founded in 2010-2019 (398 checked) and 2.2% of those founded before 2010 (273 checked).

   Source: TAM Graph · checked 2026-09-24 · 251 / 398 / 273 · method

3. 30

   18.4% of company websites in a random sample of 983 records from TAM Graph's People Data Labs company file no longer resolve in DNS.

   Source: TAM Graph · checked 2026-09-24 · 983 companies with a website · method

4. 31

   In a random sample of People Data Labs company records that TAM Graph checked in September 2026, 22.3% of companies with 1-10 employees had a domain that no longer resolves in DNS, against 12.6% of companies with 11-50.

   Source: TAM Graph · checked 2026-09-24 · 660 companies with 1-10 employees; 183 with 11-50 · method

5. 32

   38% of webpages that existed in 2013 were no longer accessible a decade later, in a Pew Research Center sample of Common Crawl pages.

   Source: Pew Research Center, 2024 ^[11]

6. 33

   A quarter (25%) of all the pages Pew collected from 2013 through 2023 were no longer accessible as of October 2023.

   Source: Pew Research Center, 2024 ^[11]

7. 34

   23% of news webpages and 21% of government webpages contain at least one broken link, Pew found.

   Source: Pew Research Center, 2024 ^[11]

8. 35

   54% of English-language Wikipedia pages have at least one reference link that points to a page that no longer exists.

   Source: Pew Research Center, 2024 ^[11]

9. 36

   Ahrefs found that 66.5% of the links pointing to 2,062,173 sampled websites since January 2013 had rotted.

   Source: Ahrefs, 2024 ^[12]

10. 37

    In the same Ahrefs study, 47.7% of lost links were lost because the page carrying them was dropped.

    Source: Ahrefs, 2024 ^[12]

11. 38

    A PLOS ONE study of more than one million web references in 3.5 million science, technology and medicine articles found that one in five articles suffers from reference rot, and seven in ten of the articles that cite web resources.

    Source: PLOS ONE, 2014 ^[13]

How much CRM data is wrong

1. 39

   76% of CRM users in Validity's 2025 survey said less than half of their organization's CRM data is accurate and complete.

   Source: Validity, 2025 ^[14]

2. 40

   37% of organizations lose revenue as a direct result of poor data quality, according to Validity's 2025 CRM survey.

   Source: Validity, 2025 ^[14]

3. 41

   In Validity's 2025 survey, 1 in 4 companies reported a drop of 20% or more in annual revenue tied to bad data.

   Source: Validity, 2025 ^[14]

4. 42

   45% of companies' CRM data is not prepared for AI, Validity's 2025 survey found.

   Source: Validity, 2025 ^[14]

5. 43

   37% of CRM staff in Validity's 2025 survey said they regularly fabricate data to tell leaders what they want to hear.

   Source: Validity, 2025 ^[14]

6. 44

   31% of CRM administrators in Validity's 2024 survey said poor quality data costs them at least 20% of their annual revenue.

   Source: Validity, 2024 ^[15]

7. 45

   44% of respondents to Validity's 2022 survey said they lose more than 10% of annual revenue because of low-quality CRM data.

   Source: Validity, 2022 ^[16]

8. 46

   In Validity's 2026 survey of 500 marketers, 67% of C-suite respondents admitted campaign data is at times manipulated to look better to leadership, against 38% across the organisation.

   Source: Validity, 2026 ^[17]

9. 47

   Continuous, automated monitoring was the capability marketers said would most increase their confidence in CRM data (39%), ahead of one unified platform (23%) and third-party validation or enrichment (19%).

   Source: Validity, 2026 ^[17]

10. 48

    59% of organizations do not measure data quality, according to Gartner surveys.

    Source: Gartner ^[18]

What bad data costs companies

1. 49

   Poor data quality costs organizations at least $12.9 million a year on average, according to Gartner research from 2020.

   Source: Gartner, 2020 ^[18]

2. 50

   Over a quarter of organizations estimate they lose more than $5 million a year to poor data quality, and 7% report losses of $25 million or more, according to IBM.

   Source: IBM, 2025 ^[19]

3. 51

   43% of chief operations officers name data quality issues as their most significant data priority, in a 2025 IBM Institute for Business Value report.

   Source: IBM, 2025 ^[19]

4. 52

   Nearly half (45%) of business leaders name data accuracy or bias as a leading barrier to scaling AI, IBM reports.

   Source: IBM, 2025 ^[19]

Frequently asked questions

What is the average B2B data decay rate?

The most quoted figure is 22.5% a year, from HubSpot's database decay simulator, which uses MarketingSherpa's 2.1% a month. Measured email figures sit close to it: ZeroBounce found at least 23% of 11 billion addresses invalid or risky in 2025, and MailCleanup measured 23.79% decay on one list of 912,317 addresses over 12 months. Data8 puts general data decay at 30% a year.

Why do published decay rates range from 22% to 70%?

They measure different things. 22.5% is an email and contact rate compounded from 2.1% a month. The 70.3% and 70.8% figures come from business-card surveys that count any changed field, such as a new job title or phone number, not records that became unusable. IVRIS Tech also shows that the same 3.6% monthly rate becomes 35.6% or 43.2% a year depending on the formula.

How fast do email addresses go bad?

In MailCleanup's 12-month study, 9.30% of addresses had decayed after six months and 23.79% after a year, and the pace rose from 2.79% in the first quarter to 11.91% in the last. Deleted mailboxes caused 55.31% of that decay. RevenueBase measured 3.6% business email decay in November 2024 alone.

How often do business contacts change jobs?

Median tenure with the current employer was 3.9 years for US wage and salary workers in January 2024, and 22% had been in their job for a year or less, according to the Bureau of Labor Statistics. A job change usually ends the person's work email address and changes their title and phone number.

How many company websites stop working?

In TAM Graph's DNS check, 5.1% of domains of companies marked active in Crunchbase records no longer resolved in September 2026, and 18.4% of websites in a sample of People Data Labs company records were dead. Pew Research Center found that 38% of webpages that existed in 2013 were gone a decade later.

What does bad CRM data cost?

Gartner puts the average cost of poor data quality at $12.9 million a year per organization. In Validity's 2025 survey of 602 CRM users, 37% said they lose revenue as a direct result of data quality and 76% said less than half of their CRM data is accurate and complete.

Sources

1. [1]HubSpot.↑1 ↑2
2. [2]IVRIS Tech. 2026.↑3 ↑4 ↑5
3. [3]RevenueBase. 2024.↑6
4. [4]Data8.↑7 ↑8
5. [5]ZeroBounce. 2026.↑10 ↑11 ↑12 ↑13
6. [6]MailCleanup. 2026.↑14 ↑15 ↑16 ↑17 ↑18 ↑19
7. [7]bioRxiv. 2019.↑20 ↑21 ↑22
8. [8]U.S. Bureau of Labor Statistics. 2024.↑23 ↑24 ↑25
9. [9]IndustrySelect.↑26
10. [10]IndustrySelect. “Measuring the high cost of bad contact data”.↑27
11. [11]Pew Research Center. 2024.↑32 ↑33 ↑34 ↑35
12. [12]Ahrefs. 2024.↑36 ↑37
13. [13]PLOS ONE. 2014.↑38
14. [14]Validity. “State of CRM Data Management in 2025”. 2025.↑39 ↑40 ↑41 ↑42 ↑43
15. [15]Validity. “State of CRM Data Management in 2024”. 2024.↑44
16. [16]Validity. “State of CRM Data Health in 2022”. 2022.↑45
17. [17]Validity. “State of CRM Data Management in 2026”. 2026.↑46 ↑47
18. [18]Gartner.↑48 ↑49
19. [19]IBM. 2025.↑50 ↑51 ↑52

Method

We searched Google for b2b data decay statistics, contact data decay rate and data decay statistics, read the pages that ranked, and counted how many of them repeat each number. Each outside number was traced to the company, survey or paper that produced it and checked against that page on September 24, 2026. Most decay rates on this page are vendor figures: HubSpot's 22.5% comes from a 2012 MarketingSherpa benchmark, and IndustrySelect's business-card figure is an informal seminar survey, so read them as estimates. We found two research papers that measure address or link decay (a MEDLINE email study, posted on bioRxiv and later published in the Journal of the Association for Information Science and Technology, and a PLOS ONE study of reference rot) and none that measure B2B contact decay directly. TAM Graph numbers come from random samples of its Crunchbase, People Data Labs and email records, with each domain looked up in DNS through two public resolvers; lookups that timed out were left out. A domain that resolves can still have a closed company or a dead mailbox behind it, so these are floors for decay.

TAM Graph numbers new on this page:

• 5.1% of company domains marked active in Crunchbase records no longer resolve in DNS, in a random sample of 998 that TAM Graph checked in September 2026. Random sample of Crunchbase companies with operating_status active and a domain, from 1,723,352 records TAM Graph loaded between August 2024 and September 2026 (68% of them before 2026). Dead = NXDOMAIN, or no A, AAAA or NS record, on both 8.8.8.8 and 1.1.1.1. Lookups that timed out on both resolvers were left out.
• In a random sample of 998 active-marked Crunchbase company domains that TAM Graph checked in September 2026, 9.6% of companies founded in 2020 or later no longer resolved in DNS (251 checked), against 4.3% of those founded in 2010-2019 (398 checked) and 2.2% of those founded before 2010 (273 checked). Same sample and DNS test as above, split by the founded_on year in the Crunchbase record.
• 18.4% of company websites in a random sample of 983 records from TAM Graph's People Data Labs company file no longer resolve in DNS. Random sample of the People Data Labs free company dataset held in TAM Graph (mostly companies with 1-10 employees). Same DNS test as above.
• In a random sample of People Data Labs company records that TAM Graph checked in September 2026, 22.3% of companies with 1-10 employees had a domain that no longer resolves in DNS, against 12.6% of companies with 11-50. Same sample and DNS test as the People Data Labs number above, split by the size band in the record.
• 0.4% of work email domains in a random sample of 1,000 addresses TAM Graph found since October 2025 no longer have a mail server or website record in DNS. Random sample of the emails collection. Dead = no MX and no A record for the domain on both 8.8.8.8 and 1.1.1.1. Mailbox-level checks (SMTP) were not run, so this counts dead domains only.

All 52 statistics as a CSV: b2b-data-decay-statistics.csv.

Cite this page

TAM Graph, “B2B Data Decay 2026 Statistics (Updated September)”, updated September 2026, https://tamgraph.com/research/b2b-data-decay-statistics

To cite one number, link its anchor: each statistic has one, for example https://tamgraph.com/research/b2b-data-decay-statistics#stat-1.