Your Field Reports Are Written From Memory — and Everyone Knows It

Ask any field rep when they write their call reports and you will get one of two answers. Friday afternoon, or Sunday evening.

By then a week has happened. Eleven visits have blurred into a general impression. The rep is not being lazy — they are being human, and they are reconstructing rather than recording. What lands in the system is a plausible summary of a week, not an account of eleven conversations.

Everyone in the chain knows this. The rep knows it. The manager reading it on Monday knows it. It gets filed anyway, because the alternative — asking a rep to fill in a form in a hospital car park between appointments — is worse.

What reconstruction actually costs you

The problem is not that the reports are dishonest. It is that reconstruction destroys exactly the details that would have been worth having.

Objections lose their specificity. In the room, a doctor said the reimbursement paperwork for the indication is unclear and their pharmacy pushes back. On Friday that becomes "price concerns". Price concerns is unactionable. The reimbursement paperwork is a problem you could solve, and if you saw it written the same way by four reps in three regions, you would know it was worth solving centrally.

Commitments lose their edges. "He was positive" is what survives. "He said he would try it on two patients after the September rota change" is what you needed, because that one has a date attached and can be followed up.

Timing collapses. A week written on Friday is a week without sequence. You cannot tell whether the objection came before or after the sample discussion, and that ordering is usually the whole story of the visit.

Silence looks like coverage. This is the expensive one. A reconstructed week rarely records what did not happen. An account that has not been visited in nine weeks does not raise its hand in a Friday summary. You discover it when the order stops — which is one quarter too late to do anything cheap about it.

Why the standard fix does not work

The usual answer is a better form. More fields, mandatory entries, a validation rule that will not let the rep submit without selecting an objection category.

This reliably makes things worse, for a simple reason: every field you add is paid for in the car park. The rep has four minutes between appointments. A form with fourteen fields does not get fourteen fields of truth; it gets four fields of truth and ten of whatever passes validation fastest. Mandatory dropdowns are especially good at producing this, because "Other" is always one click away.

The second usual answer is enforcement — a compliance rate on the dashboard, a conversation in the one-to-one. This produces compliance. It does not produce accuracy, and it costs you something real: the rep learns that the system is something done to them rather than something that helps them. After that, nothing you add to it gets used properly.

What a trustworthy record actually requires

Three things, and only three.

1. It has to be captured at the time, not at the end of the week. Not "the same day" — at the time, within a few minutes of walking out. Everything else follows from this. Memory of a specific conversation is intact for minutes and gone by Friday.
2. It has to cost the rep almost nothing. If capture takes longer than the walk to the car, it will not survive a busy week. This is not a discipline problem to be managed; it is an arithmetic problem. Eleven visits multiplied by three minutes of admin is thirty-three minutes a day that the rep will eventually reclaim.
3. It has to be structured afterwards, not during. This is the part most systems get backwards. Asking the rep to categorise the objection while standing in a corridor is what produces the "Other" click. Let them say what happened in their own words, in their own language, and do the structuring after.

Speaking is the only capture method that satisfies all three. A rep can say what happened in about thirty seconds — "Saw Dr Bērziņš at the polyclinic, he's still worried the reimbursement paperwork will bounce, said he'd try it on two patients after the September rota" — while walking to the car, in whatever language they think in. That sentence contains the doctor, the site, the objection with its actual specificity, the commitment, and the timing. All of it survives, because none of it had to be remembered.

The test that tells you whether your record is real

Pick a doctor. Any doctor in a territory you care about. Without asking the rep, answer these from your system:

• When were they last seen, and by whom?
• What did they object to, in their words, the last three times?
• What did they commit to, and has it happened?
• Has the sentiment moved in either direction over the last quarter?
• If this rep left tomorrow, would their successor know any of this?

If the answers are in a person's head rather than in the system, you do not have a field record. You have a compliance artefact — a thing that gets filled in because it must be, which is a different object entirely and cannot be used for the decisions you actually need to make.

That last question is the one worth sitting with. In a small distributor, reps are the relationship. When one leaves, the account history usually leaves with them, and the replacement starts from zero on customers the company has been serving for years.

What this is not

None of this is an argument for more oversight. It is nearly the opposite.

The reason managers chase reports is that they cannot see anything. The chasing is a symptom. When the record is trustworthy, the Monday conversation stops being "what did you do last week" — a question that puts a rep on the defensive and produces defensive answers — and becomes "the reimbursement objection has come up four times this month, what do we do about it", which is a question a rep is glad to be asked.

That change is worth more than the reporting itself. A rep who is not being audited will tell you things. A rep who is being audited tells you what closes the ticket.