Not a pricing page. The real numbers on custom CRM, the point where it beats a subscription, and what a build taught me that a brochure never would.
Custom CRM development means building a sales and customer system around how your business actually works, rather than bending your business around a product sold to a hundred thousand others. In the UK a mid-tier build runs £8,000 to £25,000, and the reason to consider one is rarely the build cost in isolation. It is the maths against a subscription you would otherwise pay every month for the life of the company. I want to give you those numbers, then tell you what building one actually taught us, because the pricing pages never do.
These numbers come from having done it, not from a rate card. We built a WhatsApp-first deal manager for a Dubai estate agency: leads landing in WhatsApp, qualified and routed, deals moving through a pipeline the team could see, follow-ups that did not depend on someone remembering. It is a CRM in every sense that matters, just shaped around the channel the business already lived in. That build is where most of what follows comes from.
A custom CRM is not competing against zero. It is competing against a per-seat subscription that never stops. A mid-tier Salesforce or HubSpot seat runs £75 to £90 per user each month, and that scales straight up with headcount. Ten seats on HubSpot Professional works out at roughly £54,000 over five years, against about £18,600 for a mid-tier custom build over the same period. The build feels expensive in year one and cheap by year three, because its cost is mostly one-off while the subscription compounds.
The break-even is not a fixed number, but a useful rule holds: custom tends to win within two to three years once you pass roughly 20 to 40 users. Below that, a subscription is usually the right call, and a supplier worth trusting will tell you so. Above it, or when your process is genuinely unusual, the ledger tips toward building. The bespoke software buyer's guide covers the wider build-versus-buy logic if you want the general version.
The first brief is always too big, and a CRM is the worst offender because everyone has a feature they are sure they need. We ignored most of it. Stripped down, a CRM is a contact database with activity tracking. Everything else is a choice about what to add and when. We built the ten things the team would touch every day: the WhatsApp intake, the pipeline view, activity logging, follow-up prompts, roles, and a reporting screen that answered the two questions the owner actually asked. That shipped. The wish list waited.
That sequencing is the single biggest cost lever on a CRM build. Each extra integration, each bespoke report, each unusual permission rule adds engineering time. Build the daily-use core first, put it in real hands, and let what people actually reach for decide the next thing you build. It is cheaper, and the software ends up used rather than admired.
Three things surprised us, and they are the parts worth planning for. First, the data you migrate in is dirtier than anyone admits, and cleaning it is real work that belongs in the budget, not a footnote. Second, the integrations are where the time goes: a pipeline is quick, but wiring the CRM to the tools around it, the messaging, the accounting, the calendar, is where weeks quietly disappear. Third, maintenance is not optional. Budget 10% to 20% of the build cost a year for upkeep, and treat the team that built it staying on to support it as a feature, not a cost.
The build discipline that kept it honest is the same one we run our own companies on: map the work into steps, watch the risky ones, and keep a person on the outputs that matter. There is a plain account of what that looks like day to day in a normal Tuesday, and the UK bespoke software guide covers the contract terms, ownership and day rates that surround a build like this.
If you have a small team, a standard sales process, and no unusual requirement, a subscription CRM will serve you better and I will say so. Building to save money on five seats is a false economy. The case for custom is real when the per-seat cost has grown into a tax, when a standard product forces your team into daily workarounds, or when the way you work, like a business that runs on WhatsApp rather than email, is the thing a generic tool would flatten. That last one is why the estate agency build worked. The tool fit the business instead of the other way around.
A genuinely simple single-pipeline CRM can come in at £2,000 to £5,000. A mid-tier build for a real team usually runs £8,000 to £25,000, and integration-heavy or SaaS-grade systems can reach £30,000 to £50,000 or more.
When per-seat fees stop making sense. A mid-tier Salesforce or HubSpot seat runs about £75 to £90 per user each month, so for a team of any size the subscription compounds. Custom typically wins within two to three years once you pass roughly 20 to 40 users, because the build cost is largely one-off.
Contact and company management, a pipeline you can drag deals through, activity logging, email integration, reporting, and user roles. Start with the ten things your team would touch every day and build those first, rather than the full wish list.
You should own both outright, with the source code handed over on completion and no per-seat fees. That ownership, and the freedom to change the system on your own schedule, is most of the reason to build rather than rent.
If you are staring at a rising subscription bill and wondering whether a build would pay off, that is a twenty-minute conversation, not a research project. Bring your seat count and your process and we will do the maths honestly. See how we work for the shape of it, or send the numbers via contact.
Tell us your seat count and how your team really works, and we will tell you plainly whether a custom CRM pays off or whether a subscription still wins.
Book a 30-minute fit call