What a custom CRM costs to build in the UK, the four cases where it beats Salesforce or HubSpot, and the ones where buying off the shelf wins.
A custom CRM in the UK costs roughly £40,000 to £120,000 to build and £6,000 to £18,000 a year to maintain, plus hosting. A licensed platform costs less on day one and more over five years — but the money is rarely what should decide it, and most companies asking this question are better off buying off the shelf.
Here is a case where the circumstances justified a custom build. A global luxury goods brand we have worked with had no unified view of its customers across European markets. Each country office had, over years, independently adopted its own e-commerce platform and its own point-of-sale tooling, and staff were tracking customer relationships in spreadsheets in the absence of any real CRM. A regional headquarters had already built a custom solution for its home market, but it was not reusable elsewhere. The result was duplicate customer records, no visibility of a customer’s history across boutique, web and repair-service touchpoints, and frontline staff with reporting needs an enterprise ERP system could not realistically serve.
The team had already evaluated enterprise CRM platforms and rejected them — not on capability, but on fit: licensing costs scaled with features they would never use, and off-the-shelf platforms move on their own release timeline, not the business’s. The organisation preferred to pay more up front for a system it could shape than to commit to ongoing licensing for one it could not.
What a licensed CRM actually costs
Before comparing anything, it is worth being precise about the other column, because per-seat pricing is where most of these comparisons go wrong:
Microsoft Dynamics 365 Sales — Sales Professional at £50.00 and Sales Enterprise at £80.70 per user per month, paid yearly, excluding VAT (Microsoft UK pricing).
Salesforce Sales Cloud — Enterprise at $175 per user per month, billed annually, with Unlimited at $350.
HubSpot Sales Hub — Professional at $90 per seat per month billed annually, plus a one-off $1,500 onboarding fee; Enterprise at $150 per seat with a $3,500 onboarding fee (HubSpot pricing).
For a 40-person team that is £24,000 to £38,700 a year in licences alone at UK list prices. Two caveats in both directions: list is not what a company with 40 seats actually pays, because vendors discount at that size and how much is a negotiation rather than a published figure; and list is not the whole bill either, because sandboxes, storage, integration platform seats and the premium support tier are all priced separately.
The five-year arithmetic
For the same 40-person sales and service team, over five years:
| Licensed platform | Custom build | |
|---|---|---|
| Year 1 | £24,000–£38,700 licences + £30,000 configuration | £85,000 build + £2,000–£4,000 hosting |
| Years 2–5 | £24,000–£38,700 a year + £10,000 a year of changes | £12,750 a year maintenance + £2,000–£4,000 a year hosting |
| Five-year total | £190,000–£264,000 | £146,000–£156,000 |
Licence lines are UK list prices for 40 seats before discount and before VAT. Only the custom column carries a hosting line, because running the software is part of what a licence buys. The bottom of that band is a real configuration at list price — two application instances, a managed database with a standby, backups and a staging environment, around £2,100 a year; the top of the band is our estimate, allowing for larger instances, longer backup and log retention and some headroom. Neither column includes the internal product owner a build needs and a licence does not.
At 40 seats the build is still the cheaper column over five years, and the crossover falls somewhere between year two and year four, depending on which licence tier you are comparing against and how much infrastructure the build needs — not, as is often assumed, somewhere out beyond year five. That is not the same as saying build. The licensed column is a known quantity you can walk away from at any renewal; the custom column is a delivery risk, a hiring commitment and a product you now own. What the arithmetic does say is that "it will be cheaper to buy" stops being automatically true somewhere around this size.
The licensed column is also the only one that grows with headcount. At 15 seats it is a third of the size and a build almost never pays back; at 200 seats it is five times the size and nothing else in the comparison matters much. Between those two points the deciding factor is not cost but fit.
The four cases where custom genuinely wins
Your process is the product. If how you qualify, price or schedule is a competitive advantage, forcing it into someone else’s object model degrades the thing you compete on. Bending Salesforce far enough to fit is its own expensive project — with the licences still payable.
The CRM is not really a CRM. Field service scheduling, clinical case management, logistics dispatch, lending workflows. Products sold as "CRM with an industry pack" often deliver a rough fit and an implementation partner on a day rate for the rest.
Data residency or regulatory constraints. Where records must live in specific infrastructure, or be auditable in ways a vendor cannot evidence, custom is sometimes the only route that passes review.
Licence cost scales with the wrong thing. If you need 300 occasional users — drivers, technicians, part-time staff — per-seat pricing punishes you for a usage pattern the vendor did not design for.
The cases where custom loses, and people build anyway
"The vendor’s UI is ugly." A build is an expensive redesign.
"We want to own our data." You already do. Export it.
"Our sales team is unusual." Almost every sales team believes this. Most are a pipeline with custom field names.
"It will be cheaper." Only past a scale most teams never reach — and only if someone maintains it, which is a permanent cost, not a project.
What actually drives the build cost
Custom CRM development cost is not one number, and the variation between two builds of the same nominal scope comes down to five things:
| Driver | Effect |
|---|---|
| Number of integrations | The largest single factor. Every system you must stay in sync with adds weeks to the build and maintenance for as long as it exists |
| Migration from the current CRM | Rarely under two weeks; dirty data can double it |
| Permissions model | "Everyone sees everything" is cheap; regional, role-based, record-level rules are not |
| Reporting | Ten fixed reports are trivial; a report builder is its own product |
| Mobile use | An offline-capable mobile client can add 40% to the whole build |
The middle route most people miss
You do not have to choose the whole thing. A licensed CRM for the standard parts — contacts, pipeline, email — with a custom application for the process that is genuinely yours, the two talking over an API. You pay licences only for the users who need them, and you build only the part that would never have fitted. It is less satisfying than a single system, and it is usually the right answer.
Before you decide
Write down the ten things your current setup stops you doing. If eight of them are configuration, buy. If eight of them are shape — your data does not fit, your process does not fit, your users do not fit — then a custom CRM is worth costing properly.
If you get to that point, our CRM development page sets out how we build these systems, and our bespoke software development page covers how we scope and price this kind of project. Where the real problem turns out to be the system underneath rather than the CRM on top, when to rebuild and when to refactor is the more useful question. And if the honest answer is that Salesforce fits after all, that is a much cheaper conclusion to reach before the build than during it.
Alexander Bessudnov
Alexander is an accomplished Chief Product Owner with a proven track record of leading successful product development initiatives at Magora. With a passion for innovation and a keen understanding of market trends, Alexander plays a pivotal role in shaping Magora's product strategy and ensuring the delivery of cutting-edge solutions to clients.
LinkedIn