Choosing a CRM for a 5–50-staff service firm

Choose the CRM your team will actually update and whose numbers you can read without anyone compiling them; everything else is secondary. At 5–50 staff the failure mode is never a missing feature — it is a system nobody feeds and reports nobody trusts. Selection by feature list is how firms end up paying for the wrong tool twice: once in fees, once in the spreadsheet that quietly replaces it.

What does a firm this size actually need?

Four things, in order of importance:

1. Low-friction logging. Emails, calls and meetings should record themselves wherever possible. Every field a human must type is a field that will eventually be skipped.
2. A pipeline with defined stages. Stages tied to verifiable events, not moods — a deal is where the evidence says it is.
3. A reporting feed. Pipeline value, win rate, conversion by stage and revenue against target should be readable live. The CRM exists to feed the layer described in The MD Dashboard Blueprint; if the numbers still need a person and a Friday afternoon, the tool has failed at its main job.
4. A sensible three-year cost. Not the first-year cost. Per-seat pricing compounds with headcount, and the tier you need in year two is rarely the tier you trialled.

What should you ignore?

Most of the demo. Lead scoring, revenue intelligence, forecast AI, territory management — these solve coordination problems that appear somewhere past 50 salespeople, not 15 staff. A small firm buying them pays for shelfware and, worse, adds interface complexity that suppresses the one thing that matters: whether people update the system at all.

Ignore, too, the promise that the CRM fills itself with insight. It fills itself with what your processes put into it. A CRM pointed at a weak or empty database reports precisely nothing useful — the data side has to be built deliberately, which is why we treat database building as its own workstream rather than a CRM feature.

How do you run the selection?

The mechanism is a structured trial, not a bake-off of brochures:

1. Shortlist three: typically HubSpot's free tier, one mid-market tool (Pipedrive or similar), and one open-source option. The trade-offs between the hosted and owned routes are covered in open-source CRM vs HubSpot.
2. Load real deals. Ten live opportunities, real contacts, real history. Demo data hides every problem you are trying to find.
3. Run two weeks of real work. When a salesperson sends an email, then check it logged itself. When an enquiry arrives, then check a record appeared without anyone typing. When you want Monday's numbers, then check they exist without exporting to a spreadsheet.
4. Score adoption, not features. Count the days each person actually touched the system unprompted. The tool with the highest unprompted-use count wins, almost regardless of what the other tools do better.
5. Price the winner at three years and next tier. Ask the vendor what the features you used in trial cost at your headcount plus 50%.

When a tool passes the two-week test on real deals, then commit and migrate; when none passes, the problem is your process, not the market, and a fourth trial will not fix it.

Which option fits which firm?

As a rule of thumb: no technical capacity and need it this week — HubSpot free, upgraded knowingly later. Sales-led firm of 10–30 wanting simple pipeline discipline — a mid-market tool at a modest per-seat fee. Any firm with a technical person and a preference for owning its data — open-source on a small server, at a fraction of the recurring cost. All three can work; all three fail identically when nobody feeds them.

What decides success after you choose?

Adoption — and adoption is a system property, not a personality trait. If updating the CRM costs effort and returns nothing to the person doing it, entries will stop within a quarter regardless of which logo is on the login page. The design moves that fix this — automatic capture, fewer fields, automation that pays the salesperson back — are the subject of nobody updates the CRM, and they are worth reading before you sign anything, because they will shape what you buy.

---

Next step: the Growth System Audit — £450, seven days, credited against any build — maps where your growth system leaks, whether the CRM you have (or plan to buy) can tell you the truth, and what to build first.