CRM data hygiene checklist: what to check daily, weekly, monthly and quarterly

A CRM data hygiene checklist is a short list of checks, each with an owner and a cadence: daily for what each call changes, weekly for the deals going into pipeline review, monthly for field quality, quarterly for duplicates and contact data. The checks that matter most in a sales CRM are the deal facts, such as next step, close date and champion, because they only exist in conversations and decay the moment nobody types them up.

This guide gives you the full checklist in one table, then walks through each cadence, how to measure whether it is working, and where automation helps and where it makes things worse. For the short version and what Claap does about it, see CRM hygiene.

The CRM data hygiene checklist at a glance

Cadence	Check	Owner	Done when
After every call	The call is logged on the right contact and deal	Rep, or the recorder	No call from the day is missing from the CRM
After every call	Next step and its date are written	Rep, or the recorder	Every open deal has a dated next step
Weekly	Deals with no conversation in 14 days are reviewed	Sales manager	Each quiet deal is moved, closed or given a next step
Weekly	Close dates in the past are fixed	Rep	No open deal has a close date before today
Weekly	Stage matches what the buyer said	Sales manager	Late-stage deals show the evidence the stage requires
Monthly	Fill rate of key deal fields is checked	RevOps	You know which fields are empty, and why
Monthly	"Other" and placeholder values are audited	RevOps	Picklists get the values reps actually need
Monthly	Each field has one source that writes it	RevOps	No call or email is logged twice
Quarterly	Duplicates are merged	RevOps or a dedupe tool	Duplicate rate is back under your threshold
Quarterly	Contact data is refreshed	RevOps or an enrichment tool	Bounced emails and old job titles are updated
Quarterly	Unused fields and overwrite rules are reviewed	RevOps	Every field left has an owner and a rule

Why CRM data goes bad: two different problems

Most CRM hygiene advice is about contacts: duplicates, bounced emails, job titles that changed when someone switched companies. That problem is real, and enrichment and dedupe tools are built for it.

The second problem gets less attention and costs more in a sales team. The facts that decide a forecast, such as the next step, the decision date, who signs and which competitor is in the deal, are said out loud on calls. They reach the CRM only if a rep types them in afterwards. On a busy day that does not happen, or it happens on Friday from memory.

"A lot of our meetings are in person, so when it comes to actually getting a lot of the call data in there, [we're] at the whim of whether or not a rep's on the road."

Heard on a recorded call · Sales leader, UK sports-data platform

Asking reps to report in a second place makes it worse, not better. When the CRM is one more form, people fill it badly or not at all.

"They had the feeling that they were reporting in two completely different places, that felt a bit overkill. Those people hate report twice. [...] Led to the low adoption or usage of [the CRM], or the sloppy use of [the CRM]."

Heard on a recorded call · Sales leader, European outsourced-sales firm

So a checklist has to cover both: a scheduled clean-up for contact data, and a way for deal facts to reach the CRM without depending on someone's memory.

Before you start: give every field an owner

A field nobody owns is a field nobody fixes. Before the first audit, list the fields your reports and your forecast actually use, usually fewer than 20, and write down who answers for each one. "Who" can be a person, a team or a tool.

Field	Source	Owner	Overwrite rule
Job title, email, phone	Enrichment tool	RevOps	Refresh quarterly
Next step and date	The latest call	Rep, or the recorder	Always keep the latest
Objections	Every call	Rep, or the recorder	Add each new one to the list
Champion	Discovery calls	Rep	Fill only if empty
Deal stage	Rep's judgement	Rep, checked by the manager	Rep only
Loss reason	The last calls of the deal	Rep, checked by RevOps	Fill only if empty

The overwrite rule is the column teams forget. Without it, a tool that writes to the CRM can erase a value a rep typed carefully, and reps stop trusting the record.

Daily: what each call changes

Log the call on the right record

Every external call should land on the contact and the open deal it belongs to: video meetings, dialer calls and in-person meetings alike. Meetings are usually matched by attendee email and phone calls by number. Internal meetings should stay out.

Write the next step with a date

"Follow up" is not a next step. "Send the security questionnaire by 14 October, legal to review" is. A dated next step is the single field that makes the weekly review fast.

Put facts in fields, not in a note

A summary pasted into an activity note looks complete and is invisible to every report. If the competitor, the budget or the decision date only exists in a paragraph, nobody can filter on it. Write it into the field.

Weekly: before the pipeline review

Review the deals that went quiet

Sort open deals by days since the last conversation with the buyer, not since the last edit. A deal can keep its stage for weeks with nobody talking to the buyer. Fourteen days is a common threshold for a mid-market cycle; pick yours from your own sales cycle.

Fix close dates in the past

An open deal with a close date before today is either lost or mis-dated. Either way it distorts the forecast. Ask the rep for a new date and what changed.

Check the stage against what was said

Stages slip forward on optimism. For each late-stage deal, ask for the evidence the stage requires: a named decision-maker, a budget, a date the buyer gave. If the calls never covered it, the stage is a hope, not a fact.

Monthly: field quality

Measure the fill rate of your key deal fields

For the fields your forecast uses, count the share of open deals where each one is filled. A field below half on open deals is not a hygiene problem, it is a process problem: nobody is writing it at the moment the information is available.

Audit "other" and placeholder values

Required fields filled with "other", "TBD" or a single space are worse than empty ones, because they look complete. If a value keeps coming back as "other", the picklist is missing the real answer, or the field is required too early in the deal.

Check that each field has one writer

If your dialer logs calls in the CRM and your call recorder does too, every call appears twice. Decide which tool writes what, and switch the other off.

Required fields have side effects on tools too. A required field your integration cannot fill blocks the whole record:

"We're missing email addresses [...] so the data can't transfer to Salesforce because of that required field."

Heard on a recorded call · Revenue team, application-security company

Require fields at stage exit rather than at creation, and test every integration against your validation rules.

Quarterly: duplicates and contact data

Merge duplicates

Run a dedupe pass with your CRM's own duplicate management or a dedicated tool. Merge rather than delete, so the activity history of both records survives. Then look for the source: an import, a form, or an integration that creates a new company instead of finding the existing one.

Refresh contact data

People change jobs, emails bounce, companies rebrand. An enrichment pass on active accounts and open deals keeps the contact side current. See our guide to CRM enrichment for the tools and the trade-offs.

Retire unused fields and review the rules

Every field nobody fills is noise in every report and every form. Archive the ones nobody has filled in a quarter, and check that every remaining field still has an owner and an overwrite rule.

How to measure CRM hygiene

Four numbers, tracked monthly on open deals, tell you whether the checklist is working:

• Key field fill rate: the share of open deals where each forecast field is filled with a real value, not "other".
• Dated next step rate: the share of open deals with a next step dated in the future.
• Quiet deal rate: the share of open deals with no conversation with the buyer in your threshold, such as 14 days.
• Duplicate rate: the share of contacts and companies flagged as possible duplicates.

The first three are about deal facts, and they move fastest when the facts are written from the calls themselves.

Where automation helps, and where it hurts

Automation helps where a human is the bottleneck: logging the call, writing the next step, filling the fields a buyer answered out loud. A tool that reads the call and writes a typed value into the field, a date, a picklist value, a number, takes the Friday catch-up out of the process. That is what Claap does with CRM auto-complete on HubSpot and Salesforce, from video, dialer and in-person calls.

Automation hurts when it writes without rules. A sync that overwrites a value a rep typed, or creates a new company when it cannot find the existing one, makes the CRM dirtier. Before you connect any tool, ask three questions:

1. How does it find an existing contact before creating a new one?
2. Can you set, field by field, whether a new call overwrites, appends or leaves the value alone?
3. Can you review values before they sync, at least for the first weeks?

Claap answers them with a rule per field (only if empty, append when possible, always overwrite) and a switch to hold values for review. It does not dedupe or enrich contacts: keep a dedicated tool for that part. For automating the work around the CRM more broadly, see CRM automation.

A 30-minute weekly hygiene routine

1. 5 minutes: list open deals with no buyer conversation in 14 days.
2. 5 minutes: list open deals with a close date in the past or no dated next step.
3. 10 minutes: send each rep their list, and ask for a date, a move or a close.
4. 5 minutes: spot-check three late-stage deals against what the buyer said on the last call.
5. 5 minutes: note any field that keeps coming back empty, for the monthly review.

Run it before the pipeline review, not after. Clean data is what makes the review about deals instead of about the data.