The CRM Nobody Opens

Most businesses of a certain size have one. Bought with reasonable intentions, configured over a few weekends, populated enthusiastically for about six weeks, and now open in nobody's browser.

The usual explanation is discipline. People are busy, the team never got into the habit, there should have been more training.

That explanation survives because it is nobody's fault in particular. It is also almost always wrong.

Disuse is information

A CRM does not get abandoned by accident. It gets abandoned because, for the person expected to update it, the effort of updating it exceeds what they get back.

That is not laziness. That is an accurate assessment, arrived at quickly and correctly by somebody who does this work all day. If logging a call takes ninety seconds and nothing in their week improves as a result, they have measured the exchange and declined it. Every time.

The question worth asking is not how to make them comply. It is what they were being asked to give and what they were getting.

In most small operations, the answer is that the CRM was set up to produce reports for somebody who does not use it, by people who get nothing from filling it in. Adoption problems are usually that shape.

This is the same test that decides whether any automation survives: does it add a step to someone's day or remove one. Nobody demos the boring one makes the case that the second kind never needs an adoption programme, because there is nothing to adopt.

The tell

Here is the diagnostic that settles it quickly.

Ask the person who is supposed to update the CRM where they actually keep track of their work.

There is always an answer. A notebook. A spreadsheet. The inbox, sorted a particular way. Notes on their phone. Whatever it is, it is the real system, and it exists because they needed one and the official one did not qualify.

Look at what that shadow system contains and what it leaves out. It will be sparse — just the fields that person genuinely needs to do the next thing. That list is the specification for what the CRM should have asked for, and the gap between it and the twenty-three required fields in the official system is the whole story.

It is worth resisting the obvious response, which is to make the official system mandatory. Compliance can be enforced, and when it is, the logging rate goes up and the data quality does not — people enter whatever value lets them close the dialog and get back to work. You end up with a fuller database that is less trustworthy than the sparse one, and with a team that now has a reason to resent it.

Why adding AI to it makes things worse

The current instinct is to put something intelligent on top: summarize the notes, draft the follow-ups, score the pipeline.

If the underlying records are half-complete and six weeks stale, this produces confident output from bad data. Which is worse than no output, because a blank field looks empty and a generated summary looks authoritative. Somebody will make a decision on it.

There is a version of this that works, and it runs the other way round. Instead of asking the system to interpret what people entered, have it capture what already happened — the call that occurred, the email that was sent, the meeting that was booked — so that the record fills itself from activity rather than from somebody's memory at the end of a long day.

That inverts the exchange. The person stops paying a tax to feed a system and starts getting a record they did not have to build. Adoption stops being a behaviour problem because there is no behaviour to adopt.

It also changes what "our data is a mess" means. Most of the historical mess does not need cleaning, because nothing depends on it. What needs to be reliable is narrow: that a new inquiry gets a record, an owner, and a status. Three fields, applied consistently from today, are worth more than a quarter spent tidying five years of dead opportunities — and the tidying project is the one that never finishes.

The uncomfortable part

Sometimes the honest finding is that the CRM should not be there at all.

A six-person business with one pipeline and forty active opportunities may genuinely be better served by a shared spreadsheet that everyone actually updates than by a proper system that nobody does. The spreadsheet is not a failure to grow up. It is a system whose cost of use matches what the operation gets from it.

That is not usually what anyone wants to hear, because the CRM was bought as a signal of seriousness and abandoning it feels like a step backwards. It is the same trade described in can you just buy a tool for this — the right purchase is the one that matches the operation, not the one that matches the ambition. It is not. Keeping something that produces unreliable data is the step backwards, because decisions get made on it anyway.

Nobody needs to be persuaded to check something that saves them a step. The reason the CRM sits unopened is that it never did.

If you want to find the other places where that exchange is running the wrong way, the leak audit takes an afternoon and a notebook.