CRM data decay is the slow loss of accuracy in your contact records: the people, companies and email addresses in your CRM (customer relationship management system, the database where sales and marketing keep contacts and deals) keep changing in the real world while the records stay frozen. A contact changes jobs on a Tuesday, a supplier gets acquired in March, an intern imports a trade-show list with half the emails mistyped, and none of it shows up as an error. You cannot stop this, but you can slow it a lot. The four controls that work are verification at every entry point, a re-verification cadence set by segment, a named owner with a small decay dashboard, and a rule to suppress dead records rather than delete them. Here is where the rot comes from and how to build those controls without a big project.
Why CRM data decay is gradual, and why that makes it dangerous
A database does not go bad the way milk does. It goes bad the way a roof does: one loose tile at a time, until a storm shows you how many are missing. The storm is usually a big send or a quarterly pipeline review.
Nobody gets an alert saying “412 contacts became invalid this month.” The bounce rate creeps from 1% to 2% to 4% over a year, reps get used to a certain number of dead calls, and by the time someone asks why the numbers look off there is no single event to blame.
Five ways a contact database rots
People change jobs, and the record stays behind
In most B2B databases this is the biggest single mechanism. When a contact leaves, their employer usually shuts the mailbox within weeks. The CRM record does not know; it still shows a name, title, company and email that now describe nobody.
The US Bureau of Labor Statistics reported a median employee tenure of 3.9 years in January 2024, meaning half of all workers had been in their job for less time than that. If half your contacts are under four years into a role, a meaningful slice of last year’s list has already moved on, and in my experience sales, marketing and tech roles churn faster still.
Companies rename, merge or close
A company gets acquired and the old domain is retired six months later. A startup rebrands and the old addresses start bouncing. A small business closes and the domain lapses. Sometimes the domain lives on as a catch-all (a mail server that accepts every address, so a verifier cannot tell which mailboxes are real), which is worse in a way: nothing bounces, so you never learn the contact is gone.
Typos and duplicates at the point of entry
Every manual entry point leaks. A rep types “gmial.com”. A prospect fills a form on a phone and fat-fingers their own address. A second rep creates a duplicate because the search did not match a different spelling of the company. Individually tiny; across a team of ten entering contacts every day for three years, they become a structural share of the database.
Integrations importing junk
Modern CRMs sit at the end of a pipe: webinar tools, chat widgets, enrichment vendors, the marketing platform’s own sync. Every one of them can push records in, and most do so without checking whether the email is real. One misconfigured sync can create thousands of contacts overnight with empty names and role addresses like info@ and sales@ (shared inboxes that reach a department, not a person). The email side of this decay has its own mechanics; the CRM version is broader, because titles, company names, phone numbers and ownership fields rot too.
Nobody owns hygiene
Ask who owns the CRM and you will get a name. Ask who owns the accuracy of the records inside it and you will usually get a pause. Sales thinks marketing does. Marketing assumes ops does. Ops has it on a list under “someday”. A database without an owner decays at the natural rate because nobody is measuring it.
The symptoms ops teams actually see
Decay shows up in four places before anyone calls it a data problem.
Bounce spikes on sequences. A sequence is an automated series of sales emails sent over days or weeks. When a rep enrols 200 stale contacts, the hard bounces (permanent failures because the mailbox does not exist) hit your sending domain in a cluster. Mailbox providers notice, and deliverability dips for the whole team, including the rep who did nothing wrong.
Reps chasing dead contacts. Fifteen minutes researching an account, a call, a voicemail, two emails, then a colleague mentions that person left last year. Multiply by a team. Wasted selling time is the most expensive symptom and the one that never appears in a dashboard.
Reporting that overstates the pipeline. Leads per month, contacts per account, marketable audience size: all inflated by records that describe nobody. Forecasts built on them look fine until quarter end.
Per-contact billing on a bloated database. Most marketing platforms and several CRMs price by the number of stored or marketable contacts, so dead records cost real money every month. The HubSpot cleanup walkthrough shows how one platform’s billing treats them; the pattern is similar elsewhere.
Mechanism, symptom, control
| Mechanism | What you notice | Control that slows it |
|---|---|---|
| Job changes | Hard bounces on older contacts; reps told “she left” | Re-verify by segment; refresh titles on active accounts |
| Company renames, mergers, closures | Whole domains bouncing or turning catch-all at once | Domain-level checks during re-verification; flag catch-all results |
| Typos and duplicates at entry | Bounces on brand-new contacts; two records per person | Real-time verification on forms and manual entry; dedupe rules |
| Integrations importing junk | Sudden record spikes, empty names, role addresses | Verify on import; limit which integrations may create contacts |
| No owner | Everything above, slowly, with nobody alarmed | Named owner, three-metric decay dashboard, quarterly review |
How to slow CRM data decay
1. Control the entry points
The easiest bad record to fix is the one that never gets in. Put real-time email verification on every form that creates a contact, and run every import through a bulk check before the rows land in the CRM. The two modes are compared in real-time vs bulk email verification; in short, real-time catches typos while the person is still on the page, and bulk catches the file before it lands.
Ideally one verifier does both, so that “valid” means the same thing at the front door and in the periodic cleanup. Reoon Email Verifier fits that shape: a live API for forms and CRM integrations, a WordPress plugin that verifies during form submission, and CSV upload for bulk re-verification, at an advertised 99%+ accuracy. Whatever tool you pick, make sure it flags catch-all, disposable and role addresses separately; a plain bounce check misses all three.
2. Re-verify on a cadence set by segment
One schedule for the whole database is too aggressive for the dormant half and too lax for the pipeline. Split it:
- Hot pipeline and active sequences: monthly. A dead address here costs you reputation immediately.
- General marketable database: quarterly. Often enough to catch job changes before they accumulate into a bounce spike.
- Dormant records: before reuse, never on a timer. Anything untouched for a year gets verified when someone wants to email it again, and only then; checking contacts nobody plans to use is wasted spend.
The reasoning behind these intervals is in how often to re-verify your email list. The short version: a valid address has a shelf life, and it is shortest for the contacts you care about most.
3. Give the database an owner and a small decay dashboard
The owner does not need to be senior, just thirty minutes a week and a mandate. Their dashboard needs exactly three numbers:
- Hard bounce rate on the last 30 days of sends. Trend matters more than level: a steady 0.8% is fine, a climb to 1.5% over two months is decay you can act on.
- Percentage of active contacts not verified in the last 90 days. This is your exposure. If it drifts above a third of the marketable database, the cadence has slipped.
- Duplicate rate. Records sharing an email, or a name-plus-company pair, as a share of the total. Rising means an entry point is leaking.
Three numbers, reviewed monthly, with the owner allowed to pause an integration or add a form check. That is the whole governance model, and in my experience it beats a forty-page data quality policy nobody reads.
4. Suppress the rot, don’t delete it
When a contact turns up invalid, the instinct is to delete. Resist it. That dead address sits on a record that may hold deal history, call notes, a signed contract and a relationship a rep spent a year building. Suppress the address (stop sending to it, and mark it so nobody re-enrols it) and keep the record. If the person resurfaces at a new company, you still know who they are.
Archive rather than delete for the same reason. Move confirmed-dead contacts into an archived state with a reason code: “invalid email”, “left company”, “domain retired”, “duplicate of record 4471”. When someone later asks why a contact vanished from a list, the answer is right there. The Mailchimp cleanup guide shows how archiving works on a marketing platform; most CRMs offer an equivalent through a status field or lifecycle stage.
Where I would stop being strict
Honest opinion: if your CRM has under five thousand contacts and one person doing sales, most of this is overkill. Verify your forms, run a bulk check twice a year, and get back to selling. The dashboard, the cadence and the reason codes earn their keep once there are several reps, several integrations, and a marketing team paying per contact. Below that, the one discipline that matters is refusing to import a list you have not verified.
Start with a measurement
Export a sample of your oldest active contacts, verify it, and look at the invalid percentage. That one number tells you whether the controls above are urgent or merely sensible. To run that sample now, create a free Reoon account; it comes with free verification credits every day, no card required, enough to check a slice of your database and see the rot for yourself.
FAQ
What is CRM data decay?
CRM data decay is the gradual loss of accuracy in the contact and company records in your CRM. The real world keeps changing while records stay fixed: people change jobs, companies rename or close, and new entries arrive with typos or as duplicates. The database looks full but describes fewer real, reachable people every month.
How fast does CRM data decay?
Faster than most teams assume, and the rate depends heavily on your audience: B2B contacts in high-turnover roles rot quicker than consumer lists. Rather than trusting a published industry percentage, measure your own. Verify a sample of contacts added a year ago and note how many are now invalid. That figure is your real decay rate and the right basis for a cadence.
Should I delete invalid contacts from my CRM?
Usually no. Suppress the email address so nobody sends to it, and archive the record with a reason code such as “left company” or “domain retired”. The deal history and notes on that record still have value, and the person may reappear at a new employer. Reserve deletion for genuine junk: test rows and gibberish form fills.
How often should I re-verify CRM contacts?
Set the cadence by segment. Contacts in active sequences or the live pipeline deserve a monthly check because a bounce there hurts immediately. The general marketable database is fine on a quarterly cycle. Verify dormant records only when someone plans to use them again, since checking contacts nobody will email is wasted effort.
