Choosing a CRM system is one of those decisions that looks simple from a distance and gets complicated up close. There are dozens of credible options, pricing that is deliberately hard to compare, vendors who all claim to do everything, and colleagues who all have opinions based on a system they used five years ago. Get it right and the business runs better for years. Get it wrong and you are either stuck with an expensive mismatch or ripping it out in eighteen months.
This guide gives you a ten-step process for choosing well. It is written for the person in a UK SME who has been handed this job — often the sales director, operations manager or MD — and needs a method rather than a recommendation.
Step 1: Define the problem before the product
Write down, in plain sentences, what is broken or missing today. "We lose track of follow-ups." "Nobody knows the real pipeline number." "Customer information lives in four places." This list is your anchor. Every later decision — requirements, demos, scoring — refers back to it. If a vendor's shiny feature does not address something on this list, it is a distraction.
Be specific about who feels the pain. A CRM chosen to solve management's reporting problem but imposed on a sales team whose problem is admin overload will be resisted. The problem statement should reflect the users' reality, not just the buyer's.
Step 2: Map your processes
Before looking at any software, document how work actually flows: lead handling, quoting, onboarding, account management, support. Include the exceptions — the awkward cases are where implementations succeed or fail. Note volumes too: how many leads a month, how many active customers, how many users. These numbers determine which tier of system you need and what it will cost.
This step typically takes one to two weeks of conversations and produces a process map of a few pages. It is the highest-value work in the whole selection process, and most businesses skip it.
Step 3: Write requirements as user stories
Translate the process map into requirements written from the user's perspective, with acceptance criteria:
- "As a field rep, I can see a customer's full order history on my phone in under a minute."
- "As a sales manager, I can see every open opportunity over £10k with no activity in 14 days."
- "As the MD, I get a weekly pipeline summary by email without asking anyone."
Mark each as must-have, should-have or nice-to-have. Be ruthless: most businesses have 10–15 genuine must-haves. Everything else is negotiable. This list becomes your demo script and your scoring sheet — it stops vendors steering the conversation toward their strengths and away from your needs.
Step 4: Shortlist three, not ten
The CRM market has a long tail, but for most UK SMEs the realistic shortlist comes from a familiar set: HubSpot, Salesforce, Pipedrive, Zoho, Monday, Capsule, and sector-specific options. Pick three that plausibly fit your size, complexity and budget. Considering more than three wastes everyone's time; considering one is not a selection process.
If you are still running the commercial side of the business on spreadsheets, read our CRM vs spreadsheets comparison first — it covers the point at which spreadsheets stop coping and what changes when you move.
A rough sizing guide: under 10 users with straightforward sales, the simpler tools (Pipedrive, Capsule, HubSpot Starter) are usually enough. Ten to fifty users with complex processes, integrations or reporting needs, look at HubSpot Professional, Salesforce or Zoho. Beyond that, or with genuinely unusual requirements, add a custom-built CRM to the comparison — sometimes building exactly what you need costs less over five years than licensing and heavily customising a platform that fights you.
Step 5: Run scripted demos
Do not accept the vendor's standard demo. Send each vendor your ten most important user stories a week in advance and ask them to demonstrate each one, live, in the product. Score each story: works out of the box, works with configuration, needs custom development, cannot do it.
Scripted demos are revealing. Vendors who can do your must-haves natively will show you gladly. Vendors who cannot will try to steer back to their standard tour — which tells you what you need to know. Take notes in the scoring sheet during the demo, not after; memories of demos blur together within days.
Step 6: Check the total cost of ownership
The licence price on the website is the beginning of the cost, not the end. Build a three-year TCO for each option including:
- Licences for your actual user count, at the tier you actually need (not the advertised starter tier that lacks the features in your must-haves).
- Implementation: configuration, data migration, integration with your other systems, training.
- Add-ons you will need: common gotchas include advanced reporting, extra storage, API access, and the marketing tools that are priced separately.
- Price rises: check the contract terms. Some vendors raise prices significantly at renewal once you are embedded.
- Internal time: the staff hours for requirements, testing, training and ongoing administration. This is real cost even though it does not appear on an invoice.
Our breakdown of the real cost of CRM in the UK covers the items businesses most often miss. As a rule of thumb, first-year implementation typically costs one to three times the annual licence fee.
Step 7: Ask the questions vendors hope you won't
In the final evaluation, ask directly:
- Show me a customer of our size, in our sector, using this today — and let me speak to them without you in the room.
- What happens to our data if we leave? In what format, how quickly, and what does extraction cost?
- Which of our must-haves require custom development, and what does that cost to build and maintain?
- What are the API limits, and what do integrations actually cost to build against your platform?
- How has your pricing changed in the last three years, and what are the renewal terms?
- What is your product roadmap for the next 12 months — and what are you deprecating?
The quality of the answers matters as much as the content. Evasion on data extraction or pricing is a red flag you should take seriously.
Step 8: Pilot before you commit
Before signing a multi-year deal, run a pilot: one team, real data, four to six weeks, with success criteria agreed in advance. A pilot surfaces the issues demos hide — data migration pain, workflow friction, the features that looked good in a demo and annoy in daily use. It also gives you leverage in contract negotiation: you are a proven customer, not a prospect.
Negotiate the contract after the pilot, not before. Push for a shorter initial term with renewal options rather than a three-year lock-in. Get the price-rise cap in writing.
Step 9: Plan data migration early
Data migration is where timelines slip. Start it during evaluation, not after signing: audit what you have, decide what migrates and what gets archived, cleanse before importing. Anything untouched in two years probably should not come across. Assign a named person to own each dataset. Test the migration with a sample before doing the full load.
Step 10: Budget for adoption, not just software
The selection process ends when you sign. Success depends on what happens next: training on real workflows, a phased rollout, weekly adoption tracking, and an owner with authority. Ring-fence time and budget for the first 90 days after go-live — this is when the investment either pays off or quietly dies.
If you want help with any part of this — requirements, vendor evaluation, or implementation — get in touch. We are a CRM development company in Manchester, and we will give you a straight answer, including when the right answer is an off-the-shelf product rather than anything we would build.
FAQs
How long does it take to choose a CRM system?
For an SME, allow 6–10 weeks: two for process mapping and requirements, two for shortlisting and scripted demos, two for TCO analysis and reference checks, plus a 4–6 week pilot before committing. Rushing this decision is how businesses end up with expensive mismatches.
What is the biggest mistake when choosing a CRM?
Choosing based on feature lists and vendor demos rather than your own documented requirements. The second biggest is comparing licence prices instead of three-year total cost of ownership, which hides implementation, add-ons, price rises and internal time.
Should a small business buy Salesforce?
Not necessarily. Salesforce is powerful but complex and expensive to implement well; many businesses under 20 users get better value from simpler tools like Pipedrive, Capsule or HubSpot Starter. Match the system to your actual complexity, not your aspirations.
How many CRM vendors should we evaluate?
Three. More wastes time; fewer is not a real evaluation. Send all three the same scripted demo based on your must-have user stories and score them on the same sheet.
Is it better to build a custom CRM or buy off-the-shelf?
Buy when your processes are fairly standard; build when your requirements are genuinely distinctive or when heavy customisation of a platform would cost more over five years than building exactly what you need. Our comparison of custom vs off-the-shelf software walks through the trade-offs.
