Most small businesses buy a CRM for the wrong reason: because it feels like what a "proper" business does, not because the spreadsheet actually failed. The spreadsheet usually didn't fail — it just got messy, which is a five-minute fix, not a £200-a-month subscription.
Quick answer: You need a CRM once deals are genuinely being lost to poor tracking — no one's sure who last contacted a prospect, follow-ups slip because nothing prompts them, or two people chase the same lead without knowing it. Short of that, a tidy spreadsheet or a disciplined inbox does the job for free. Try a free-tier CRM before paying for one, and only consider a custom build once an off-the-shelf tool genuinely can't hold your pipeline's shape.
The three stages, and what each one costs
| Stage | What it is | Rough cost | Best when |
|---|---|---|---|
| Spreadsheet or inbox | Deals tracked in a shared sheet or your own memory of email threads | £0 | One person selling, low deal volume, nothing time-sensitive |
| Off-the-shelf CRM | HubSpot, Pipedrive, Zoho or similar, free or paid tier | £0-£450/month for a small team | Multiple people touch leads, or follow-ups need a prompt to happen at all |
| Custom-built CRM | Software shaped around your exact pipeline and other systems | £6k-£15k+ | Your sales process genuinely doesn't fit any standard CRM's stages, or it must integrate tightly with something bespoke |
Skipping straight to a paid CRM before the spreadsheet has actually failed is the most common overspend in this category — it's rarely the tool that was missing, it's the discipline to use it.
How do you know you've outgrown the spreadsheet?
Answer: When deals stop moving and nobody notices until the prospect has gone quiet — the sheet stops reflecting reality, not just looking untidy.
The signs, in the order they usually show up:
- Follow-ups depend on memory. Nothing prompts anyone to chase a lead after three days of silence, so some just don't get chased.
- More than one person can touch the same lead. Two people email the same prospect, or a handover between colleagues loses context because nothing structured was handed over.
- The pipeline stage is a guess. "Is that one still live?" gets answered by scrolling back through email rather than reading a status.
- Deals go cold with no post-mortem. A lead disappears and nobody can say when it stalled or why, so the same mistake repeats.
One of these is a bad week. Two or more, consistently, means the spreadsheet has stopped doing the one job a sales tracker has to do.
Should you try a free CRM before paying for one?
Answer: Yes — a free tier proves the discipline is real before you pay for seats, and most small teams can run on it indefinitely.
HubSpot's free CRM, Pipedrive's trial, or a well-structured Airtable base with stages and a reminder automation all do the same core job: force a deal into a stage, and prompt a follow-up when it stalls. Run it for a full sales cycle. If the team actually updates it and follow-ups genuinely happen on time, you've proven the process — and you may never need to pay for anything more than the free tier, or a low per-seat plan once the team outgrows it.
The exception is a business that already knows its pipeline is unusual — multi-stage approval chains, several products with different sales motions, or a need to sit tightly inside a system already in use. In that case, a generic CRM trial won't tell you much either way; that's a scoping conversation, not a free-trial test.
When is a custom-built CRM actually worth it?
Answer: When an off-the-shelf CRM's stages, fields, or integrations genuinely can't model how your business actually sells — not because the default setup feels slightly wrong.
Score the business against these:
- The pipeline shape is genuinely non-standard — multiple parallel approval tracks, or stages that don't map to any CRM's default "lead → qualified → proposal → won" flow.
- Deep integration is load-bearing — the CRM needs to read and write live data from a bespoke system (stock, scheduling, a proprietary quoting tool) that no off-the-shelf connector handles well.
- Per-seat pricing has become the actual cost driver — a large sales team on a premium tier can make a one-off custom build cheaper over 2-3 years than an ongoing subscription.
- You've genuinely tried and outgrown at least one off-the-shelf tool — not assumed you would, tried it.
Two or more true, and a custom build earns a proper look. Fewer than that, an off-the-shelf CRM — even a mid-tier paid one — is almost always the cheaper, faster, lower-risk choice.
A worked example: an installer business drowning in a shared inbox
A 12-person heating and plumbing installer quotes jobs from a shared inbox — enquiries come in by email or phone, get logged in a notebook by whoever answers, and quotes go out from memory of "roughly what we did for a job like that."
Step 1 — is deal volume actually low? No. Around 40 enquiries a month, several people answering the phone, no shared view of what's been quoted or chased.
Step 2 — worth trying a free CRM first? Yes, and worth doing before anything else — the pipeline here is a completely standard "enquiry → quoted → won/lost" shape, exactly what any off-the-shelf tool models out of the box.
Step 3 — genuinely non-standard, deeply integrated, or seat-cost-driven? No. Nothing here needs custom fields beyond what Pipedrive or HubSpot ship with by default.
The maths: two jobs a month are currently being quoted late enough that the customer books a competitor instead — call it £1,800 a month in lost work, conservatively. A paid CRM tier for a 4-person sales/office team runs roughly £150-£300/month. The payback is immediate, and it doesn't need a custom build to get there.
Verdict: this one's a clear off-the-shelf CRM case, not a custom build — the mistake to avoid here isn't under-investing, it's over-investing in something bespoke this pipeline doesn't need.
When should you NOT get a CRM at all?
- You're the only one selling. A CRM adds structure for coordination between people. With one person, a spreadsheet with a "next action" column does the same job for free.
- Deal volume is genuinely low. A handful of enquiries a month doesn't need a system — it needs someone to reply promptly, which no software fixes.
- You haven't tried fixing the spreadsheet first. Adding a "last contacted" and "next follow-up" column often solves 80% of the problem before any new tool is involved.
- Nobody will actually use it. A CRM that the team doesn't update is worse than a spreadsheet — it looks authoritative and lies. Adoption is the real risk in this category, not the software.
What does it actually cost?
- Off-the-shelf CRM, small team — HubSpot, Pipedrive, Zoho: roughly £0 (free tier) to £90/user/month, so £75-£450/month for a 5-person team on a paid tier.
- Custom-built CRM — tailored pipeline, bespoke integrations: roughly £6k-£15k to build, plus hosting.
- A CRM nobody uses — the real cost isn't the subscription, it's the deals it quietly fails to save while everyone assumes it's handling things. Our full pricing guide covers where this sits against other custom-software spend.
How to actually start
Add two columns to your existing sales sheet — "last contacted" and "next follow-up date" — and see if that alone fixes the problem. If it doesn't, run a free CRM tier for one full sales cycle before paying for anything. Only look at a custom build once you can point to a specific way a standard CRM's shape doesn't fit your pipeline, not a feeling that it might.
If you want a straight read on whether your current setup is actually costing you deals, our AI System Audit covers sales and ops tooling as part of a fixed-scope review — a written report, not a sales pitch for a CRM build you may not need.