Choosing a CRM for a small business: four questions that decide it

Almost every article about choosing a CRM for a small business is a ranking, and a ranking is the wrong shape for this decision. The tools are far more alike than the comparison tables suggest. All of them store a contact, a company, a deal and a note against each. All of them sync your mail. All of them have a pipeline board.

What separates a CRM that is still the system of record in two years from one that quietly turns back into a spreadsheet is not on the feature list. It is four decisions, and the vendor makes none of them.

What a CRM is actually for

A CRM answers one question: what is true about this customer, and which system says so.

Everything else — the board, the sequences, the dashboards — is downstream of that. When a CRM fails in a small business it is almost never because the software could not do something. It is because two people wrote the same fact in two places, nobody decided which one wins, and after a few months the team stopped trusting the record and went back to their own sheets. The tool then gets blamed and replaced, and the new one fails the same way about nine months later.

That is the failure worth designing against. The four questions below are how.

1. Which system owns which field?

Write down, before you buy anything, which system is allowed to change each fact you care about. Not which systems have the field — most of them will — but which one is right when they disagree.

The usual answer for a small business:

You can pick a different split. What you cannot do is skip the exercise, because the alternative is that every system quietly claims all of it, and the first time your CRM total and your invoiced total disagree you will have no rule to resolve it with — which is the problem behind almost every reporting argument.

This document is worth more than the platform choice. It is also, conveniently, the thing you can write before you have decided which CRM to buy.

2. What happens at the seam?

A CRM is never alone. There is a form somewhere, a mailbox, an invoicing tool, probably an ad platform, and in a growing business a support inbox. Every place two of those touch is a seam, and seams are where records go missing.

So the question to ask about a shortlist is not "does it integrate with X". They all claim to. It is: when the sync fails at 2am, does anyone find out? Most small-business CRM integrations fail silently, and a silent failure is worse than no integration, because the record looks complete and is not.

If the answer is that nobody would find out, the integration needs to be something you can see the state of — which is a different piece of work from installing it, and the piece most teams skip. It is also most of what connecting a CRM to the systems around it actually involves.

3. Who maintains it after the person who set it up leaves?

A CRM configured by one enthusiastic person and understood by nobody else is a liability with a monthly fee. Small teams do this constantly: custom fields nobody can explain, three pipelines where one is used, an automation that fires on a stage that was renamed.

Two practical tests before you commit to a platform:

• Could somebody who is not technical add a field and understand what breaks? Some platforms make this genuinely easy. Others make it easy to do and hard to undo, which is worse.
• Is the configuration readable? You want to be able to open a settings page and see what the automations are, not discover them by watching things happen.

The honest version of this question is that the platform matters less than deciding, now, who holds it. If the answer is nobody, the CRM will be as good as the day it was set up and no better, and no platform on any list fixes that.

4. What does leaving look like?

Ask this before you arrive, not when you want to go. Every CRM will export contacts. The things that do not come out cleanly are the ones you will miss: conversation history, custom field definitions, file attachments, and the mapping between a deal and the invoice it became.

You are not planning to leave. You are checking that the data is genuinely yours, which is a reasonable thing to establish while you are still a prospect and they are still being helpful. If a migration ever does happen, the difference between a fortnight and a quarter is almost entirely how well the outgoing system exports.

Where the platforms actually differ

With the four questions answered, the shortlist mostly picks itself. The differences that survive contact with a real small business are these:

Every one of those is a decent system. The engagements that go badly are never the ones that picked the wrong logo. They are the ones where nobody answered question one.

What to do this week

Start the definitions document. One page, five rows, the fields your business argues about. Take it to the free trial of whichever two platforms you are between, and try to break your own rules in each of them.

You will learn more in an afternoon of that than in a week of comparison tables, and you will finish with the artefact that outlives whichever tool you choose.

Frequently asked questions

Does a small business actually need a CRM?

If more than one person talks to customers, or if you are following anything up more than a week later, yes. Below that, a well-kept spreadsheet is honestly fine — and it is better to run a spreadsheet you trust than a CRM nobody updates.

How much configuration should we do before going live?

Less than you want to. Set the fields in your definitions document, one pipeline, and the owner rule. Add nothing else until a real gap shows up in use. Most of the wreckage we are called in to clear up is configuration added in month one for a process that never happened.

Do we need the integrations on day one?

Connect the mailbox and the calendar immediately — those are what make the record complete without anyone typing. Everything else can wait until you have seen how the team actually works. An integration built on a guess about the process is a guess you then have to maintain.

The numbers in our CRM and our accounting system do not match. Is that the CRM's fault?

Almost never. It is usually two different definitions of the same word — when revenue counts, what a closed deal is, whether tax is included. That is a decision to be made and written down, not a bug to be fixed, and it is the most common thing we get called about.