Nearly every growing tech company I work with has a CRM, and a surprising number of them are not really using it. The licence is paid. The data is three months stale. The real pipeline lives in someone's head, a spreadsheet, and a channel in Slack.
The instinct is to blame discipline. It is almost never discipline.
Why it died
CRMs usually get set up in one of two ways, and both kill adoption.
The first is the full build. Someone configures everything the tool can do, because it is all there and it all looks useful. Forty fields on a deal. Custom objects. Six pipelines. Now logging a call takes four minutes, and nobody has four minutes.
The second is the reporting build. The fields exist so that a founder or a board can run a report. Every one of them is a tax on the person entering it and a benefit to somebody else.
If the only person who benefits from a field is you, the person filling it in will stop.
The smallest CRM that works
Strip it back until every field either helps the seller do their next action, or is genuinely load-bearing for a decision you actually make.
On a deal, that is usually about eight things:
- Company and primary contact
- Stage - evidence-based, describing what the buyer has done
- Value - and be honest about whether it is a guess
- Expected decision date - the buyer's date, not your hope
- Next action and its date - the single most valuable field in any CRM
- Source - so you learn what works
- Who else is involved - the second and third contact on their side
- The problem in the buyer's words - one free-text line, quoted
That last one earns its place every time. It makes the record useful to a human picking the deal up cold, which is the actual test of a CRM.
The rule that fixes hygiene
One rule, and it replaces most policies: every open deal has a next action with a date, or it is not open.
It works because it is enforceable at a glance, it is self-serving for the seller - the CRM becomes their to-do list rather than a filing cabinet - and it exposes stalled deals automatically. A deal with no next action is a deal nobody has a plan for.
Automate the parts nobody enjoys
Most of what people hate about CRMs is retyping. Connect the email and calendar so activity logs itself. Auto-create the contact from the form. Let the tool set the source. Every field a human does not have to touch is a field that stays accurate.
Be careful with the automations that create records, though. I have seen an integration quietly generate hundreds of deal-shaped records from social connections, which looked like a full pipeline and was noise. Automation should reduce typing, not manufacture optimism.
How to restart one that is already dead
Do not migrate the mess. Do this instead:
- Export what is there and keep it somewhere safe. You will not need it as often as you fear.
- Delete or archive every deal that fails the next-action test. The number will drop hard. That is the point.
- Cut the fields to the list above, then add back only what someone asks for twice.
- Re-enter only live deals by hand. It takes an afternoon and it makes the data yours.
- Run one weekly review off the CRM and nowhere else. If the meeting uses a spreadsheet, the CRM is dead again by month two.
That last step is the one that decides it. A CRM is adopted when it is the only place the important conversation happens, and not before.
The tool matters less than you think
HubSpot, Zoho, Salesforce, Pipedrive - at your stage they will all do this. The difference between a working CRM and a dead one is almost never the software. It is whether the person entering the data gets something back.
So open yours and count the fields on a deal. How many of them help the person filling them in?
Want to know where your sales engine is leaking?
The free Growth Scorecard takes about six minutes and shows you where growth is breaking - pitch, pipeline, process, people or tools. If you want the full write-up plus a prioritised 90-day plan built with me, that is the Growth Roadmap.
Get your free scorecard See how we work together