DiagnosticCRM & data18 Sept 20266 min

Is your CRM telling you the truth?

The system went in. The training happened. Two cycles later the activity report and the field tell different stories, and nobody can say which one is real.

Who this is for
Commercial leaders, sales force effectiveness and CRM owners. Most useful if you are about to re-tier accounts, redesign a call plan, or sign off a number built on field-reported activity.

Every commercial team I have worked with has a CRM. Most have a version of the same problem. Compliance reads in the high eighties on the slide, and everyone plans as though the data were not there. Nobody says so out loud, because the number looks fine.

Reviews ask whether the field is using the system. Wrong question. What matters is whether the record is something someone saw, or something someone remembered. Write up the morning after and you are reporting. Write up fourteen calls on the twenty-ninth and you are remembering, under pressure, against a target. Both go in green. Only one of them can carry a re-tiering.

A compliance number tells you the form was filled in. It tells you nothing about when.
CYCLE CLOSEDAY 1DAY 30
Read the shapeActivity records created by day of month, one incentive cycle. The monthly volume is respectable. The distribution is the diagnostic, and it is the chart almost nobody runs.

Five checks follow. None of them needs a data extract, a vendor or a steering committee, and each wrong answer points at a different failure. The fixes are not interchangeable, which is why so many of them miss.

The diagnostic

Five questions, in this order

1

Ask three reps where they keep the notes they actually read before a specialist call.

Not where they log. Where they look.

If it is a notebook, a phone note or a shared sheet The CRM is a reporting layer, not a working tool. Reach and frequency, call plan adherence and tier assignment are all sitting on top of a record nobody consults before the meeting that matters.
2

Plot record creation by day of month, and against the incentive cycle close.

One query. You are reading the shape, not the volume.

If entry spikes in the final three days You are looking at recall written against a target rather than observation. Treat the cycle's frequency data as unusable rather than merely imperfect, and do not re-tier off it.
3

Ask a first line manager what they did differently last week because of something they saw in the system.

One specific instance. Not what the dashboard could show. What changed.

If they cannot name one The system is an audit instrument and the field knows it. The real management system is the conversation in the car, which leaves no record and does not scale.
4

Count the mandatory fields on a call record, then complete one on a phone between appointments.

Time it honestly.

Over six fields, or over ninety seconds You are taxing the behaviour you are trying to buy. Every additional required field trades a little truth for a little structure, and the trade is rarely priced.
5

Ask what happens to a rep who does not update. Then ask what happens to one who does.

The second answer is the diagnostic.

If the first is a reprimand and the second is nothing The record costs the field time and returns them nothing. What you are seeing is a rational response to the incentive you built, not a discipline problem.

Reading the pattern

Fail on one and two and the problem is the instrument. That is a design job, and a solvable one.

Fail on three and five and the instrument is probably fine. You have a management problem: slower to fix, less comfortable to name, and far cheaper than the re-implementation most organisations buy instead. The usual sequence is to assume a tooling failure, spend eighteen months and a seven figure budget proving it was not, and arrive at the management conversation anyway.

Fail on four and you have bought structure at the cost of truth, which is the most common and least discussed trade in field force design.

One caveat worth stating. This assumes a field force that is broadly willing and reasonably led. Where that is not true, every question above will read as a tooling failure when it is something else entirely, and no amount of system design will touch it.
One page

Take it into the room

The same diagnostic on a single A4 sheet, sized to print and hand across a table.

Download the PDF →

Run it and tell me what you found

If two or more came back wrong and you want a view on which fix is worth spending money on, that is the kind of thing I am useful for.

Tell me what needs to change →