// Guides · Adoption

Why don't sales reps update the CRM, and what actually fixes it?

By Kevin Bluett, co-founder of Ahoy · Updated September 2, 2026

Reps don't update the CRM because the update gives them nothing back. Logging a call serves the manager's report and the forecast, not the next deal, so it loses to the next call every time. Cutting required fields, running reviews from the record, and paying for hygiene each raise compliance a little, and none of them hold. The fix that lasts takes logging out of the job: capture activity from email, calendar, and calls automatically, and ask the rep only for judgment.

Why don't reps update the CRM?

The CRM asks and does not give. A rep who logs a call gets a cleaner report for their manager and a more confident forecast for the company, and nothing at all for the next conversation with that buyer, which is the only thing the rep is measured on. So the update sits below the follow-up and the proposal, and it stays there.

The fields ask for things the rep does not know yet. Close date and amount are required on most deal records at creation, and on a first call the rep knows neither, so they guess. The guess becomes the forecast. Nobody goes back to correct it, because that is another update with no payoff, and by then the number has already been rolled up into a forecast somebody presented.

The record is stale by the time they open it. If the last update was two weeks ago, catching up means reconstructing two weeks of email and meetings from memory, which is archaeology rather than selling. The longer the gap, the more it costs to close, so the gap grows.

The manager reads the report and the rep never does. The CRM's outputs are dashboards, pipeline views, and forecast rollups. The rep works out of the inbox and the calendar, and maybe a notes app. When the people reading a system and the people writing to it are different people, the writers stop writing. I spent about a decade leading CRM engineering at HubSpot and shipped plenty of required fields myself, and it took me most of that time to see this.

The top performers are the worst updaters. In the free audits we run, the best rep on the team is nearly always the one with the emptiest records. The rep with the fullest calendar has the least time to log it and the least fear of the consequences, and chasing them for updates takes selling time from the person you least want to slow down.

One founder described his old CRM to us on a call as "data entry, not a source of insight." This summer we watched someone ask an AI assistant which AI CRM could ingest their Gmail and calendar data, because "our founders still sell half the deals and forget to log activities." That question is the whole problem in one sentence: the people closing the most business are the ones the record knows least about.

The scale is not in dispute. Salesforce's 2026 State of Sales research puts the share of a rep's week that goes to non-selling work at about 60 percent, and names manually entering customer notes into the CRM as one of the tasks eating it (Salesforce, State of Sales, seventh edition). A large share of that 60 percent is the CRM update, and it is the share most teams try to fix by asking for more of it.

What does "CRM adoption" actually measure?

Most adoption dashboards count logins, records created, and fields completed. I shipped a few of those dashboards too, and I have changed my mind about what they measure: whether reps are touching the tool, not whether the tool knows what is happening in the pipeline. A rep can log in every day and update nothing, and a list import can create hundreds of contacts nobody has spoken to.

Two numbers tell you what you want to know.

Capture rate is the share of buyer-facing activity that lands on the right record. Over a week, count every email to or from a buyer, every meeting with one, and every call, then count how many of those appear on the correct contact or deal in the CRM and divide the second number by the first. If that number is low, the CRM is not recording much of anything, whatever the login chart says.

Deals with a dated next step is the share of open deals that carry a specific next action with a date attached. Count the open deals, then the ones whose next step holds a real action and a future date rather than "follow up" or a blank, and divide. It tells you how much of the pipeline is being worked rather than remembered.

Metric	What it tells you	What it misses
Weekly active users	Whether reps open the tool	Whether they changed anything, or read anything, once inside
Records created	Volume of contacts and deals added	Whether those records are real, current, or duplicated from an import
Fields completed	How many required fields hold a value	Whether the value is true or a placeholder typed to close the form
Capture rate	How much of the real buyer activity the CRM knows about	Whether the rep has formed a view on what happens next
Deals with a dated next step	How much of the pipeline is actively being worked	Whether the buyer has agreed to that step; check inbound replies for that

The first three rows are effort metrics and the last two are truth metrics. Effort rises when you nag; truth rises when the system knows more. In my experience the first kind is what gets reported upward, because it is what the tool makes easy to chart.

Which fields actually matter?

Run a required-fields audit before anything else. List every field that must be filled to save a deal, then ask of each one whether a report you run today depends on it. Not a report you might build. One you run this week.

Five fields survive that question on almost every team: stage, next step with a date, close date, amount, primary contact. Everything else can be optional until a report needs it, and when one does, make the field required then and write the report's name on it.

Every extra required field raises the cost of the update and lowers the truth of the data at the same time. Each required field is a question the rep has to answer before saving, and each one is a reason to close the tab and do it later. And when a rep must fill a field they cannot answer, they fill it with a placeholder: "TBD" in the competitor field, a round number in the amount, the last day of the quarter as the close date. Those placeholders show up in the forecast looking like facts. Fewer required fields means fewer placeholders, and the data you do collect is more likely to be real.

What fixes work, and how much?

Teams try the same handful of fixes. Here is how each holds up, and for how long:

Run pipeline reviews from the CRM record only. If the manager reviews only what is on the record and refuses a spreadsheet or a Slack message, updates happen the night before the review. This works, and it stops working the week the manager travels or lets one exception through, because the reps learn the rule is soft.

Cut required fields. Works, permanently, and the effect is small, because removing friction from an update that still gives the rep nothing back moves compliance from low to slightly less low. Do it anyway. It is free and it improves the truth of what you collect.

Make the CRM the rep's own tool. Put the rep's next actions, reminders, and today's meetings in it, so they open it for their own reasons. This works only when the CRM knows enough about their deals to say something worth reading, and a CRM that was built to remember rarely does.

Incentives and leaderboards. A spiff for data hygiene works for about a month. After that the reps who care have done it, the reps who don't have priced it in, and the leaderboard mostly measures who games it best.

Mandatory training. No. Reps know how to update the CRM; they are choosing not to, for the reasons above. This is the one I'd push back on hardest, because it treats a decision as a skills gap.

The structural fix is different in kind. Instead of raising compliance with a task the rep does not want, remove the task. Email, calendar, and call systems already hold nearly every activity the CRM wants logged, so a tool that reads those sources and writes the record itself makes the update happen without anyone deciding to do it. What is left for the rep is judgment: which stage, what the buyer really said, whether to push or wait.

A version a team can run this month:

Audit required fields. List every mandatory field on the contact and deal records. Keep stage, next step and date, close date, amount, and primary contact. Make the rest optional. Write down which report each remaining required field feeds.
Connect email and calendar capture. Turn on whatever automatic logging your CRM supports for the inbox and calendar, or add a tool that does it. Every buyer-facing email and meeting should land on the right contact and deal with no rep action. Sample ten deals a week later and count what got missed.
Switch the manager's report to capture rate. Stop reporting logins and record counts. Report capture rate and the share of deals with a dated next step, weekly. When the number the manager reads is the number the system produces, nagging stops being the tool.
Stop asking reps for what the system already knows. Drop every pipeline-review question the captured record can answer: when did we last hear from them, who was on the call, what was sent. Ask the rep only what the record cannot know: what happens next, and why.

A worked example, as an illustration rather than a benchmark; your numbers will differ. Say a team of six reps, each with around 40 buyer-facing emails and three meetings a day, which is about 240 emails and 18 meetings a day for the team, call it 1,300 activities a week. Logged by hand, the CRM sees perhaps a third of that. Meetings usually get a note, the important emails sometimes do, and the rest never do, so call it 430 activities on the record out of 1,300, a capture rate around a third. Your own number comes from the sample in step two.

Now connect capture. Every email to or from a buyer and every calendar event lands on the right record, and calls come through the dialer or the meeting tool. Capture moves from around a third to nearly everything, with the misses being new contacts not yet matched to a deal, and nobody on the team typed anything to get there.

The second number moves for a different reason. Before, a dated next step depended on the rep typing one after every call, so perhaps half of the team's 60 open deals had one. After capture, the system reads the last meeting and the last thread and drafts the next step itself, "Send security questionnaire by Thursday," dated, with the follow-up email already written, and the rep approves it in one tap or edits it. Once the next step is drafted for the rep rather than typed by them, the share of deals with one stops being a compliance question and becomes a review question: did the rep agree? On this team it goes from about half to nearly all, and the Monday sweep gets shorter because every deal has a date to sort by.

When is adoption not the real problem?

Sometimes the CRM is empty because the pipeline is empty, and no amount of capture will fill it. Before spending a quarter on adoption, check four things.

Is there enough activity to capture? If a rep has three buyer conversations a week, the CRM is not the bottleneck. Prospecting is.
Do the stages mean anything? If nobody can say what has to be true for a deal to enter Proposal, reps are not skipping the update; they are guessing at a question with no answer. Fix the definitions first.
Does the manager use the record? If pipeline reviews run from a spreadsheet or a Slack thread, the reps have correctly concluded that the CRM is optional. The manager's habit is the adoption problem.
Is the data model fighting the business? A relationship-driven firm forced into a deals-only model, or a group with three entities sharing one pipeline, will avoid the tool because the tool is wrong, not because the reps are.

If all four check out and the record is still empty, it is an adoption problem, and the rest of this guide applies.

What to look for in a tool

The criteria are short, because the problem is specific:

Does it capture inbound and outbound email, calendar events, and calls onto the right contact and deal without the rep doing anything?
When it sees a meeting end or a thread go unanswered, does it draft the next step and the follow-up for approval, or does it add a task to the pile?
Does it keep a human in the loop? One-tap approval on a drafted email is a guardrail. Autonomous sending is a different product with a different risk profile; decide on that separately.
Does it show the rep their own deals and next actions first, so the CRM is useful to them and not only to the manager?

Ahoy is built around capture. It writes every email, call, and meeting to the right record itself, then drafts the next step and the follow-up for one-tap approval, so the only thing a rep types is a decision. If you want to measure your own capture rate first, the free CRM audit is a 45-minute look at how much of your pipeline activity reaches the record today, with no data connected and nothing to install.

Frequently asked questions

What is CRM data entry?

CRM data entry is the manual work of recording sales activity in the CRM: creating contacts, logging emails and calls, updating deal stages, and filling required fields. It is the part of the CRM that reps avoid, because it takes time and gives nothing back to the deal. In a capture-first system most of it disappears, because the record is written from email, calendar, and call data rather than typed.

How do you automate CRM data entry?

Connect the sources where the activity already lives: the inbox, the calendar, and the calling or meeting tool. A capture-first CRM reads those and writes emails, meetings, and calls onto the right contact and deal on its own. The remaining manual work is judgment: confirming a stage change, approving a drafted next step, correcting a mismatched contact. Automate the record and keep the human on the decisions.

How do you measure CRM user adoption?

Not by logins or record counts. Measure capture rate, the share of buyer-facing emails, meetings, and calls that appear on the right record, and the share of open deals with a specific next step and a date. Both can be checked by sampling ten deals a week. They rise when the system knows more, not when reps are nagged more.

What is the best approach to CRM adoption?

Stop treating it as a compliance problem. Cut required fields to the five that reports need, run pipeline reviews from the record, and then remove logging from the rep's job by capturing activity automatically. Ask reps only for what the system cannot know: what they think happens next and why. Adoption follows when the CRM gives the rep something back.

Which AI-based CRMs cut data entry?

Look for capture-first tools: ones that read email, calendar, and calls and write the record themselves, then draft the next step for approval instead of adding a task. Ahoy is one example, built to act rather than to remember. Others exist, and the category is described in our guide to AI-native CRMs. Whatever you evaluate, ask to see a deal record after a week with no rep input.

Book a Demo Get a Free CRM Audit