How to Get Sales Reps to Actually Use the CRM

You do not get reps to use the CRM by asking harder. Every lever most teams reach for — mandatory fields, adoption dashboards, tying it to comp — reliably makes the data worse, because none of them address the actual reason. Here is what does.

By Dan Pearson · 9 min read · CRM

You do not get sales reps to use the CRM by asking harder. Every lever most teams reach for — mandatory fields, adoption dashboards, naming and shaming in the pipeline meeting, tying CRM hygiene to commission — reliably makes the data worse rather than better. Not because reps are lazy, but because none of those levers touch the actual reason the CRM is empty.

I spent fifteen years on the other side of this. I have been the rep filling in the fields on Friday afternoon, and I know exactly what I typed when I could not remember what happened on a call from Tuesday. It was not a lie. It was a reconstruction, which is worse, because it looks like data.

The actual reason

Updating the CRM is unpaid work that happens after the paid work is done.

That is the whole problem in one sentence. A rep's day is calls, demos, follow-ups, negotiations — the things they are measured and compensated on. CRM entry is a tax levied on top of it, payable in the tired part of the day, in a system that gives them nothing back.

Every rep does this arithmetic, consciously or not: what do I get for the next twenty minutes of typing? And the honest answer, in most organisations, is "my manager gets a report." That is not a motivation problem. That is a correctly-solved arithmetic problem.

So the question is not how to make reps comply. It is how to change the arithmetic.

Why the usual fixes backfire

Mandatory fields. The intuition is that if the field must be filled, it will be filled with something true. What actually happens is that it gets filled with "N/A", "TBD", a single space, or whatever value clears validation fastest. This is worse than blank, because a blank field is honest about being empty while "TBD" looks like data to every system reading it downstream. If you have ever sorted a required field by most common value and found junk in the top three, you have seen this.

Adoption dashboards. Measuring logins and record-touches produces logins and record-touches. Reps will open records and close them. You have created a metric, not information.

Tying hygiene to comp. This one works, in the narrow sense that the fields do get filled. It also teaches the team that the CRM is a compliance instrument rather than a tool, and it quietly incentivises optimistic stage-setting at exactly the moment you most need honesty. You will get better-looking data and a worse forecast.

The training refresh. Reps rarely fail to update the CRM because they cannot find the button.

Notice what all four have in common: they attack the symptom while leaving the arithmetic untouched. The rep still pays the tax; you have just raised the penalty for non-payment.

What actually works

1. Make capture automatic rather than obligatory

The single highest-leverage change: stop asking a human to record what happened, and let the system observe it. Call recording with transcription and structured summary. Meeting notes drafted from the actual conversation. Email threads attached to the right deal without anyone dragging anything.

The rep's job shifts from writing to confirming — twenty minutes becomes twenty seconds. That is not a productivity improvement, it is a change of category. This is rung 2 on the Pipeline Automation Ladder, and it is the rung that fixes adoption as a side effect of fixing something else.

2. Remove fields, do not add them

Go through every required field and ask what decision it actually informs. Most fail the test. Fields survive in CRMs for years past the death of the report that justified them, and each one adds friction to every single record.

Cutting required fields feels like giving up on data quality. It reliably improves it, because the remaining fields get real answers.

3. Give the rep something back

The CRM should return value to the person feeding it, not just to their manager. Concretely: the pre-call brief that assembles itself from the last three interactions. The follow-up that drafts itself after the call. The alert that tells the rep a deal has gone quiet before the forecast meeting does.

The moment a rep opens the CRM because it helps them close, adoption stops being a project.

4. Make stage changes evidence-based

Instead of asking a rep to assert that a deal is at Proposal, let the system propose the stage from what actually happened, and have the rep confirm or correct it. You get two things: a stage you can trust, and a record of why it moved, which is what makes a forecast auditable three months later when nobody remembers.

5. Fix the worst path first, publicly

Do not run a "CRM improvement initiative." Pick the single most-hated piece of manual entry — usually call logging — automate it, and let the team feel it. Credibility for everything else comes from that first thing working.

What to measure instead of adoption

Three numbers, none of which a rep can game by opening records:

Adoption rate measures whether people complied. These measure whether the system is earning its place.

A 30-day sequence

Week 1 — measure. Ask reps how many hours a week they spend on CRM admin, then pull a sample of 20 recent stage changes and see how many you can explain from the record alone. Both answers will be worse than you expect. Write them down; they are your baseline.

Week 2 — subtract. Remove every required field that does not inform a decision someone actually makes. Announce it. This buys goodwill you will spend later.

Week 3 — automate the worst path. Usually call and meeting capture. Recording, transcription, structured summary written back to the right record, rep confirms in seconds.

Week 4 — give something back. Turn on one thing that helps the rep directly: a pre-call brief, a drafted follow-up, a stalled-deal nudge that arrives before the pipeline review rather than during it.

Then re-measure. If seller hours have not moved, you automated the wrong path — which is useful information, and cheaper to discover in a month than in a year.

Frequently asked questions

Our reps say the CRM is fine and they just forget. Is this still the problem? Usually, yes. "I forget" is what people say about work that has no immediate payoff. Nobody forgets to submit an expense claim.

We are on HubSpot and it has AI features. Do we need more? Possibly not. Most teams have native capture features they have never configured properly. Turn on what you already own before buying anything — the audit checklist covers what to look for.

What if leadership insists on the adoption dashboard? Keep it, and add capture coverage next to it. When automated capture lands, the two diverge visibly — adoption stays flat while coverage jumps — and that gap is the most persuasive argument available to you.

Does this work with a team that has already given up on the CRM? It works better, because expectations are on the floor. The first automation that removes real work is disproportionately convincing to a team that has been asked to try harder several times.

Related: why AI CRM projects fail — the five failure modes, including the compliance spiral this article describes.