Getting a CRM going without making it a project

Start with one question it has to answer

Before you type anything in, write down the question you want to be able to ask every morning. For most businesses that's *who do I need to ring today, and what did they say last time?* Sometimes it's *which quotes are still out?* or *which customer haven't I spoken to in six months?*

That one question is your measure. Every setting, every field and every stage you think of afterwards gets judged against it: does this help answer the question, or is it something that might be interesting one day? Leave the second kind out. You can always add it later; what you can't undo is the feeling that the system is a lot of work.

Week one: only what's live

Don't put your whole contact list in. Put in where something is actually happening: the quotes that are out, the conversations from the past month, the customers you expect something from this month. For a firm of a few people that's usually between ten and forty lines, and it's half an hour's work.

The gain isn't only time. Your old list is full of companies that have stopped trading, contacts who work somewhere else now and addresses from three moves ago. Migrating means moving the mess with you. Adding someone the next time you speak to them means your list is right from day one.

Agree exactly one rule

This is the most important step in the whole piece, and it costs nothing. Agree one rule with yourself and your colleagues: after every conversation, three lines and a next step with a date on it.

Three lines, not a write-up. What they wanted, what you said, what's needed. And then a task, with a date, even if that date is "in two months". Without that second half you're building an archive rather than a system: pleasant to read back through, but it never does anything on its own.

If there's more than one of you, agree who owns a relationship as well. Not out of hierarchy, but because a task with no owner gets done by nobody.

Set your stages to your real process

Almost every CRM arrives with a default pipeline. Change it to match what actually happens at your place, rather than what sounds tidiest. If you always go and look at the job before you can price it, then "visited" is a stage. If your quotes have to go through a procurement team, that's a stage.

Keep it to between five and seven steps. More stages feel more precise but produce uncertainty: with twelve steps nobody is sure when something moves from seven to eight, and at that point the split means nothing. There's more on how to build one in the piece on setting up a pipeline.

After four weeks: check that it's being kept up

Put a slot in your diary now, four weeks out. You're checking one thing: does every live opportunity have a next step with a date on it? Not whether the system looks nicely filled in, but whether it's still alive.

If the answer is no, it's almost always one of three things:

1. Filling it in takes too many steps. Strip fields out until only the ones you actually read back are left.
2. You're not opening it at the right moment. It should be the first thing you open in the morning, not something you tidy up at the end of the week.
3. Two systems are running side by side. That's the most common cause, and the only fix is to get rid of one, even if that one was better.

Four mistakes you can skip

Wanting it all at once. Contacts, pipeline, mailings, integrations and reporting in week one isn't a rollout, it's a refit. Take the one question from the start, and widen out once that works.

Mandatory fields. They look as though they raise quality; in practice someone types "n/a" and your data is worse than it was. Make the name mandatory and nothing else.

Reporting before there's anything to report on. A dashboard built on three weeks of use shows noise. Look at numbers once there's a quarter in there; until then the only useful measurement is whether it's being kept up.

Not naming anyone. Someone has to be the person who looks once a week at whether deals are standing still. In a firm of five that isn't a job title, but it does need a name attached.

What a realistic plan looks like

Half an hour to set up an account and enter the work that's live. A quarter of an hour to adjust your stages. One five-minute conversation with your colleagues to agree that single rule. And then four weeks of simply working.

If a supplier is selling a firm of five people a six-week implementation — "onboarding", usually — what you're mostly buying is the complexity of their product. The SME Digital Adoption Taskforce, reporting to the Department for Business and Trade, found that smaller firms hold back from software precisely because taking it on looks too hard and switching feels too risky. A rollout long enough to need a plan is evidence for them, not against.

In Nermio this is the reason nothing needs configuring before you start: the six stages, the roles and the day view are already there. You put one company in and look again in four weeks to see whether you're still keeping it up.