CRM data audit checklist

Most CRM audits end with a vague sense that the data is bad and a cleanup project nobody finishes. The problem is that 'bad data' is not a finding. This checklist turns the audit into a set of queries with a number, a threshold, and an owner, so the result is a list of fixes instead of a mood.

Category:: CRM & Data Hygiene
Formats:: PDF + CSV + web view
Sections:: 6
Updated:: September 24, 2026

What you get

A scorecard of 12 measurable checks, each with the query logic and a starting-point pass threshold
Duplicate detection by matching key (email, domain, normalized company name) rather than by eyeball
Stage integrity checks that catch deals the forecast is counting but should not be
A findings log that assigns every failed check to an owner with a fix date
A prose guide on which failures to fix at the source versus clean up in bulk

Who it's for

RevOps and CRM admins who inherited a CRM they did not build
Sales ops teams preparing for a forecast process change or a migration
Leaders who suspect the dashboard is wrong but cannot say where

What's inside

1
Audit scorecard
7 columns, 5 worked example rows
2
Completeness and ownership checks
10-point checklist
3
Stage integrity checks
6-point checklist
4
Duplicate matching keys
5 columns, 3 worked example rows
5
Findings log
7 columns, 1 worked example rows
6
Fix the source, then the records
Guidance notes

Preview of section 1

Audit scorecard

Thresholds are starting points. Replace them with your own targets after the first run. Result is the percentage your report returns; Pass is Y or N.

Check	Object	How to measure	Starting threshold	Result	Pass	Owner
Open opps with close date in the past	Opportunity	Open opps where close date < today, divided by all open opps	Under 5%			Sales ops
Open opps with no activity in 30 days	Opportunity	Open opps with last activity date older than 30 days, divided by all open opps	Under 15%			Sales managers

The preview shows part of section 1. The full template has all 6 sections (5 not previewed here), with blank rows ready to fill in. Download the full template

How to use it

1
Run the numbers before you open a single record
Build each check as a saved report or list view first. The audit is the report output, not a sampling exercise. If you start by clicking into records you will fix twenty and learn nothing about the other twenty thousand.
2
Set your own thresholds after the first run
The thresholds in the scorecard are starting points, not benchmarks. Run the audit once, record where you actually are, then set a target you can hit next quarter. A threshold nobody can meet gets ignored.
3
Trace every failure to the process that caused it
A 30% blank industry rate is not a data problem, it is a form or an enrichment problem. For every failed check, write down where the bad record entered the CRM. Fixing records without fixing the entry point guarantees you run this same audit again with the same result.
4
Assign owners by object, not by audit
Sales ops owns opportunities, marketing ops owns leads and contacts, CS owns renewal records. RevOps runs the audit but should not own every fix, or the fixes stop when RevOps gets busy.
5
Rerun quarterly on the same date
Put the audit on the calendar for the second week of each quarter, after the prior quarter's closed deals settle. Trend lines across four audits are worth more than any single score.

Frequently asked questions

What should a CRM data audit include?

Completeness of the fields your reports depend on, duplicate rates by a defined matching key, ownership (records owned by inactive users), and stage integrity (deals in the wrong stage or forecast category). Each check should be a query with a number and a threshold, not a spot check.

How often should you audit CRM data?

Quarterly for the full audit, with a handful of checks (past-due close dates, stale opportunities) run weekly as part of pipeline hygiene. Quarterly gives enough time for source fixes to show up in the numbers.

What is an acceptable CRM duplicate rate?

There is no universal standard. A common rule of thumb is to aim for low single digits on contacts and accounts, but the right target depends on your matching key and data sources. Measure your current rate first, then set a target you can reach next quarter.

Should we clean data manually or with a tool?

Use tools or bulk updates for anything rule-based (merging by email, normalizing picklists, reassigning inactive owners). Save manual review for opportunity duplicates and account hierarchies, where an automated merge can destroy pipeline history.