How to Measure CRM Data Health: 6-Metric Scorecard

Short answer

CRM data health is measured with six metrics: required-field completion, recency, duplicate rate, zombie-record rate, decision-maker coverage and close-reason completion. Score each one from 0 to 100 and take the average. Below 70, stop making decisions based on reports from your CRM.

What is CRM data health?

CRM data health is the degree to which the records in your CRM are complete, current and unique enough to support a decision. The question is not "how many records do we have?" but "how many of these records can carry a decision?"

Most companies ask the first question. 40,000 customer records, 1,200 open opportunities. The numbers look big. But if half of them are empty or dead, what you have is volume, not data.

Why measure data health at all?

What you do not measure stays an argument. "There is no data in the CRM" is a feeling; "38% of required fields are empty" is a finding. You can set a target against the second one.

The second reason is more expensive: your data health score is a leading indicator of forecast error. If your sales forecast has missed three quarters in a row, the problem is probably not the model but what goes into it. We covered that link in more detail in why CRM data quality breaks down.

Which six metrics do you measure?

Build the scorecard from six metrics. Each answers a single question, and each can be pulled as a report from your CRM or your ERP.

Completion rate: on open opportunities, how many required fields (amount, close date, source, owner) are filled?
Recency: how many open opportunities were touched in the last 14 days?
Duplicate rate: how many separate records exist for the same company?
Zombie-record rate: the share of open opportunities with no activity for more than 90 days.
Decision-maker coverage: in how many opportunities is the person who actually signs on record?
Close-reason completion: in how many closed opportunities is "why we won / why we lost" written down?

How do you calculate completion and recency?

Completion rate is the number of opportunities with the required field filled, divided by total open opportunities. If 150 of 400 open opportunities have no close date, your completion rate is 62.5%.

For recency, apply the same logic to dates: if the last activity date is more than 14 days old, that record counts as stale. Keep the two metrics apart, because they tell you different things. Completion shows what never got entered; recency shows that follow-up has stopped.

How do you find the zombie-record rate?

A zombie record is an opportunity in your CRM that has had no activity for more than 90 days and is still marked open. It enters the forecast, inflates the target, and nobody is working on it.

Sort open opportunities by last activity date and count everything older than 90 days. If 310 of 1,200 fall in that group, your zombie rate is 26%. A practical threshold: above 10%, the forecast has stopped meaning anything. The cleanup steps are in CRM cleanup: finding and closing stale deals.

How do you build the scorecard? A worked example

Convert each metric to a percentage, subtract the inverse ones (duplicates, zombies) from 100, then average. A sample B2B company:

Completion: 62% → 62 points
Recency: 55% → 55 points
Duplicates: 9% → 91 points
Zombies: 26% → 74 points
Decision-maker coverage: 41% → 41 points
Close reason: 30% → 30 points

Average: (62+55+91+74+41+30) / 6 = 58.8. That score says the company keeps records but never records why a sale happened. The two lowest metrics are exactly the two fields marketing needs for targeting.

These numbers are a worked example, not a benchmark. Produce your own in the same format, monthly.

Where do these metrics come from?

All six come out of the standard reporting screen in your CRM or your ERP; no new tool is needed. Export the opportunity list with four columns: amount, close date, last activity date and close reason. Add the company name and the duplicate rate falls out of the same file.

Decision-maker coverage is the one exception. In most CRMs that field either does not exist or is mixed into a generic "contact" field. Use a separate flag: does the opportunity have at least one person marked "decision maker", yes or no? The answer is binary, so it is easy to measure — and it is usually the metric that improves fastest.

Do the first measurement by hand. After that, repeating the report monthly is a ten-minute job.

What do you do when the score drops?

The usual reaction is to add required fields and send a reminder. Both work briefly, then the score falls back. The real problem is not that data entry is hard — it is that high-quality data is never produced in the first place. A salesperson learns something in a meeting and writes down whatever they still remember three days later.

So the point of the scorecard is not to blame the team but to show which information never gets captured at all. Three steps:

Pick the two lowest metrics and leave the rest alone for now.
Instead of making those two fields mandatory, connect them to where the data is actually born — the sales conversation.
Measure the score monthly and put it on the first slide of the sales meeting.

Frame it this way: what you got (1,200 records on hand), what you gave (selling hours lost to reporting, budget spent on wrong targets), what you will get (a forecast that holds and a team that knows why it wins).

FAQ

How often should CRM data health be measured?
Monthly is enough. More often produces noise; quarterly shows the drift too late.

What is a good data health score?
In practice, above 80 is healthy, 70-80 needs watching, and below 70 is where reports should not be trusted. Calibrate the threshold against your own history.

Does a small team need this too?
Yes, and it matters more. In a three-person team the knowledge sits in people's heads; when one leaves, the portfolio leaves with them.

Which metric matters most?
Close-reason completion. If you do not know why you win, you cannot make either sales or marketing repeatable.

Will adding required fields raise the score?
In the short term yes, but it produces fields filled with "unknown" and "other". Completion goes up; information does not.

Closync AI is an AI layer that turns sales conversations into high-quality CRM/ERP data and that data into sales decisions.

Last updated: 13 October 2026