How much does a custom CRM cost in the UK?
Nobody publishes real numbers for this, which makes it impossible to budget. Here is what drives the cost of a custom CRM, with worked examples you can sanity-check against any quote.
Search for this question and you will get a range of "£10,000 to £500,000" and a contact form. That is not a useful answer, and it is a large part of why small businesses find this so hard to budget for.
So here is a more honest attempt: what actually drives the number, worked examples at three realistic scopes, and the ongoing costs that quotes tend to leave out.
What actually drives the cost
Development cost is roughly a function of five things. Almost every price difference between two quotes traces back to one of them.
1. Number of distinct workflows
Not the number of screens, the number of genuinely different processes. "Track leads through a pipeline" is one. "Track leads, schedule installs, raise invoices, and manage engineer routes" is four, and it is roughly four times the work.
2. Integrations
Every external system you connect to adds cost, and the cost varies wildly with the quality of the other side's API. A modern, well-documented API might be two days. A legacy system with a CSV export and no documentation can be two weeks.
3. Data migration
This is the single most underestimated line item in the industry. Moving ten years of inconsistent records out of spreadsheets and an ageing system, deduplicating, reconciling and validating, regularly costs more than building the feature that uses the data. If a quote does not mention migration, it is not finished.
4. Number of user roles
One kind of user is straightforward. Admin, manager, staff, and read-only client access means permissions logic on every feature, and a testing burden that multiplies accordingly.
5. Compliance requirements
Handling health data, financial records, or anything requiring an audit trail adds real work: encryption, access logging, retention policies, and documented processes. Necessary, but not free.
Three worked examples
These are shaped like real projects rather than abstract tiers.
| Focused tool | Departmental CRM | Multi-site system | |
|---|---|---|---|
| Typical range | £6k to £15k | £18k to £45k | £50k to £120k+ |
| Workflows | 1 to 2 | 3 to 5 | 6+ |
| Users | Up to ~10 | 10 to 50 | 50+, multi-location |
| Integrations | 0 to 1 | 2 to 4 | 4+, incl. legacy |
| Data migration | Minimal | Moderate | Substantial |
| Typical timeline | 4 to 8 weeks | 3 to 5 months | 6 to 12 months, staged |
The focused tool
One painful workflow, done properly. A clinic's referral tracker, a contractor's quoting system, a distributor's stock reconciliation. Usually the right first project: it is small enough to deliver quickly and specific enough to produce a measurable saving you can point at when funding the next phase.
The departmental CRM
The common case. Leads, customers, jobs or appointments, documents, and reporting, with a couple of integrations to accounting and email. Replaces a spreadsheet-and-shared-inbox arrangement for a team of ten to fifty.
The multi-site system
Several locations, several user roles, real permission boundaries, existing systems that must keep working during the transition. This is genuinely a programme of work and should be staged so each phase ships something usable.
The costs quotes leave out
Build cost is not total cost. Budget for these from the start, because they are what determine whether the system is still working well in year three.
- Hosting and infrastructure. Typically £20 to £200/month for a system of this size. Modern platforms are cheap; this is rarely the problem.
- Maintenance and support. Plan for roughly 15 to 20% of build cost per year. This covers dependency updates, security patches, bug fixes, and small changes.
- Training and rollout. A day or two of your team's time, plus documentation. Skipping this is the most reliable way to waste the whole investment.
- Year-two changes. Your business will change. Reserve a budget for it rather than being surprised.
Comparing against SaaS honestly
The comparison people usually make is upfront cost, which flatters SaaS enormously. The comparison that matters is total cost over the life of the system at your actual headcount.
Take a mid-market CRM at £60 per user per month, a realistic figure once you add the tiers that include the features you need.
| Team size | SaaS over 5 years | Custom (build + 18%/yr) |
|---|---|---|
| 10 users | £36,000 | ~£45,000 |
| 25 users | £90,000 | ~£45,000 |
| 50 users | £180,000 | ~£70,000 |
At ten users, SaaS wins comfortably and you should buy. Somewhere around fifteen to twenty users the lines cross. Beyond that the gap widens every year, and the custom column buys you an asset you own rather than a subscription you rent.
Two caveats, in fairness to SaaS. It is available today, and its maintenance burden is genuinely someone else's problem. Those are real advantages that a spreadsheet comparison does not capture.
How to reduce the number
- Stage it. Build the one workflow costing you the most, measure the saving, then fund the next phase from it. Avoid big-bang projects.
- Keep what works. Do not rebuild accounting. Integrate with it. See custom software vs off-the-shelf for where that line sits.
- Clean your data first. Migration cost scales with mess. Tidying spreadsheets in-house before the project starts is free and saves real money.
- Be strict about version one. Every "while we're at it" feature adds cost and delays the date you start getting value.
If you have a quote in front of you and want a second opinion on whether the scope is sensible, that is a reasonable thing to ask about, no obligation attached.
Frequently asked questions
Because scope varies enormously and most quotes are given before scope is understood. A single-workflow tool and a multi-site system are both called a 'CRM' but differ by an order of magnitude in effort. Any quote given before someone has looked at your actual workflows and data is a guess.
Fixed price works well for a clearly defined first phase where the scope is genuinely settled. Day rates suit longer, evolving programmes. The risk with fixed price is that changes become contentious; the risk with day rates is that scope drifts. Staging the work into fixed-price phases usually balances both.
Hosting is typically £20 to £200 per month for a system of this size. Beyond that, budget roughly 15 to 20% of the build cost per year for maintenance, security updates, and small changes. If nobody has mentioned ongoing costs to you, ask before signing.
For typical mid-market per-seat pricing, the crossover tends to land somewhere between fifteen and twenty-five users over a five-year horizon. Below that, off-the-shelf is usually the better financial decision unless you have a specific fit or compliance reason to build.