Every field in a CRM has a cost. Someone fills it in, someone maintains it, and someone is slowed down by it. Very few earn their keep.

Here is the set that does.

The fields that change decisions

Where the enquiry came from

The most valuable field in most small business CRMs and the one most often neglected.

Without it you are guessing about marketing. With it, consistently recorded over a year, you can see which channels produce customers who actually buy, and which produce volume that never converts.

Two rules make it work. Keep the list short — six or seven options, not twenty-five. And avoid an "other" option that becomes the default, or ensure "other" requires a note.

Stage, and the next action with a date

The stage tells you where an opportunity is. The next action tells you whether anybody is doing anything about it.

Of the two, the next action matters more. An opportunity at "quoted" with no next action is not in your pipeline — it is a memory. This is what makes the difference between a CRM that recovers lost work and one that records it.

Value and expected date

Rough is fine. The purpose is to see the shape of the coming quarter, not to forecast to the pound. Precision here invites argument and adds no value.

What was actually discussed

Notes, in prose, dated. This is what makes a customer record useful to a colleague. "Chased, no reply" is worth little. "Waiting on their landlord's approval, expects an answer after the board meeting on the 14th" is worth a great deal.

Why you lost

Captured at the point of losing, from a short list: price, timing, chose a competitor, no longer needed, no decision made.

That last category is the interesting one. Businesses assume they lose to competitors and frequently lose to inaction, which is a completely different problem with a completely different remedy.

The fields to delete

Anything not used in a decision. Walk the field list and ask, for each: when did we last make a decision using this? Most implementations can lose half their fields with no loss.

Industry classifications nobody filters by. Standard industry codes are collected diligently and used never.

Duplicated contact details. A phone number in three fields is three chances to be out of date and no way to know which is right.

Free-text where a list would do. A source field typed by hand produces "Google", "google", "web search" and "internet" as four separate answers, and no report can reconcile them.

Required fields people invent values for. If everyone enters 0 or "n/a", the field is producing noise and a small daily irritation.

The test for any CRM field is whether a decision has ever been made using it. Not whether it might be interesting one day.

Capture it without typing it

Accuracy is a design problem. The less that is typed, the more accurate the data.

Website enquiries should create records directly, with the source set automatically from the form or the campaign. That is one field permanently correct and never filled in by a person.

Email should attach itself to the right record, so the history builds without effort — see connecting your CRM to your inbox.

Look up addresses from a postcode rather than typing them. Faster and more accurate.

Take company data from a register where possible, so names and numbers are correct and consistent.

Let accounting supply payment status, so nobody is updating "paid" by hand.

Keeping it clean

Data decays whether or not anybody touches it. People move on, companies rename, numbers change.

Prevent duplicates at entry. Warn when a similar record exists, rather than deduplicating annually.

Validate as it is typed. Email format, phone shape, postcode. Cheap and effective.

Review a sample monthly. Twenty records, checked properly, tells you more about your data quality than any dashboard.

Have one person responsible. Not for entering everything — for noticing when the data stops being trustworthy.

Retention and the law

Under UK GDPR you may hold personal data for as long as you have a lawful reason and no longer. In practice:

  • Set a retention period for each category — customers, past customers, unconverted enquiries — and record why.
  • Actually delete when it expires. A policy that is not applied is worse than none, because it documents an obligation you are not meeting.
  • Keep marketing consent separately, with a record of when and how it was given.
  • Be able to find and export everything about one person if they ask.

That last point is worth testing. A subject access request against a system where a person's data is spread across notes, attachments and email history is difficult, and the deadline is short.

A useful annual review

Once a year: list every field, mark those used in a decision, delete the rest. Check the source list still reflects your channels. Check retention is being applied. Sample twenty records for accuracy.

An afternoon, and it prevents the slow accumulation of fields that makes a CRM tiresome to use — the same drift covered in getting staff to use the system.

Our CRM team configures systems for UK businesses around the decisions they actually make. Start a conversation.