Find exactly where it breaks
Run these four checks in order. Each narrows the problem further.
01
Identify the decision-critical fields
List the five to ten fields that actually drive decisions in your CRM. Not every field, the fields that determine who gets called next, which deals get attention, which accounts are at risk, which leads get routed where. If a field is not driving a decision or filtering a view that someone acts on, leave it out of the audit. You are auditing operational usefulness, not data model completeness.
02
Run a completeness check
For each decision-critical field, calculate the percentage of records where it is filled in. Most CRMs can produce this as a simple report or export. Anything below 80 percent for a field that drives decisions is a quality problem worth addressing. Anything below 50 percent is a structural failure, the field either is not being filled in at intake or is not maintained after creation. Record the percentage and approximate count for each gap.
03
Run a consistency check on enumerated fields
For fields with a fixed set of values, status, source, category, stage, pull a count of every distinct value that appears. Compare against the intended values. Extra values reveal drift: manual overrides, old values never removed, or different teams using the field differently. For each unexpected value, estimate how many records it affects and whether those records represent a real pattern or just an error.
04
Prioritise by operational impact
Cross-reference the completeness and consistency findings against the decision-critical field list. The highest-priority fixes are fields that are both important for decisions and badly broken. A field that is 40 percent incomplete and drives primary lead routing is more urgent than a field that is 10 percent inconsistent and used only for historical reporting. Write down three to five specific, actionable fixes in order of operational impact. That list is the audit output.
After the audit
What to do with what you find.
The audit produces a prioritised list of specific problems. The next decision is which to fix at the source versus which to manage going forward. Most teams try to fix everything at once and end up fixing nothing properly.
What to fix, what to leave
How to act on audit findings
Fix at intake first: for each broken field, identify whether the break happens at creation or later. If it happens at creation, the field is optional and people skip it, make it required. This prevents the problem from recurring without requiring anyone to clean existing records.
Decide what to clean vs. what to flag: for existing broken records, decide which fields are worth correcting retroactively and which are worth marking as unreliable and moving forward with better intake rules. Not everything needs retroactive cleaning.
Assign ownership before fixing: before correcting any field, decide who is responsible for keeping it accurate going forward. A fix without an owner returns to its broken state within weeks as new records are created without the corrected rules being enforced.
If your team can feel the data is wrong but cannot point to specifically where, an audit is the right first step.
Bring access to the CRM and the reports people rely on most. The audit identifies what is broken, how badly, and what to fix first, before anyone writes a single line of cleanup logic.
Bring the page, report, or workflow as it is now.
We reply with the clearest next step, or an honest no.
