Almost every UK SMB has a CRM. Almost none of them are happy with it.
The CRM is usually the first piece of software a growing business buys, often before anyone has thought hard about what they actually need from it. Salesforce, HubSpot, Pipedrive, Zoho, Capsule, Insightly. The choice is rarely strategic; it is usually whatever the founder used in their last job, or whatever a salesperson sold them six months in.
Two years later, the team is fighting it. The bureau hears the same complaints in audit after audit. This post is about how to tell whether the CRM is the actual problem, or whether you have been blaming the tool for something the tool was never going to fix.
The "fighting the CRM" signal
The clearest signal that the CRM is the problem: somebody on the team has built a parallel spreadsheet to track the things the CRM should be tracking.
Once that exists, the CRM has lost its job. The spreadsheet is now the source of truth. The CRM is a partial mirror of the spreadsheet, kept in sync by hand because somebody has to log into it for compliance reasons or because the salesperson reports run off it. Half the team trusts the spreadsheet; the other half trusts the CRM; both are right; both are wrong.
If you have one of these spreadsheets, the CRM is failing. The next question is whether it is failing because of the CRM or because of how it was set up.
Five signs the tool itself is wrong
These are signs the CRM is the wrong shape for the work, and no amount of configuration is going to fix it:
1. The data model does not match how the team actually thinks about the work.
A recruitment firm tries to use a generic CRM and ends up bending "Companies", "Contacts", and "Deals" into "Clients", "Candidates", and "Placements". It works, sort of, but every report is a workaround and every new starter spends two weeks learning the translation. The problem is that the CRM was built for a different industry's vocabulary.
2. The team copies data out of it more than they put data in.
If the workflow is "log into the CRM, run a query, paste the result into a spreadsheet, work in the spreadsheet, occasionally update the CRM at the end", the CRM is acting as cold storage, not as a working tool. Whatever the team is doing in spreadsheets is the actual operation.
3. Every customer-facing email is being sent from outside the CRM.
The CRM has built-in email features. Nobody uses them. Salespeople send from Outlook because the CRM email tool is clunky. The result: half the customer history lives in the CRM and half lives in inboxes, scattered across personal email accounts. That is not a CRM problem to solve with more features; it is the wrong shape entirely.
4. The "single view of the customer" requires three tabs and a spreadsheet.
A senior account manager joining a call has to open the CRM, the project tool, the support inbox, and a shared spreadsheet to see the full picture of one customer. That is the work a CRM is supposed to do for them. If three tools fight over the answer, none of them is really doing the job.
5. New hires take three weeks to learn the CRM.
If the onboarding for a sales role includes a multi-day course on how to use a CRM that supposedly mirrors the team's actual workflow, the CRM is too far from the work. A tool fitted to the work is learnable in a day. The training overhead reveals the mismatch.
Three signs it is the setup, not the tool
Before you blame the CRM, rule out these:
1. Nobody has ever genuinely configured it.
If the CRM is running on default fields, default pipeline stages, and default reports, the problem is that the CRM is generic, not that it is wrong. The fix here is not a custom build; it is two days with a CRM specialist who knows the platform deeply and configures it for the team's actual workflow. Custom builds are an order of magnitude more expensive than configuration done well.
2. The team does not trust the data.
Sometimes the CRM is fine but the team has lost faith because of bad data. Duplicates, abandoned records, incorrect statuses from a botched migration. The fix is a data clean-up, not a new tool. Data clean-up costs less than rebuilding the system that holds the data.
3. The CRM is not the bottleneck, the process is.
The CRM is sometimes blamed for problems that are actually about how the team works. If salespeople are not logging activities, the issue may be incentives or management, not the tool. If hand-offs from sales to delivery are messy, the issue may be the hand-off process, not the CRM. A custom build will not fix any of those.
When a custom build is genuinely the right answer
A custom build replaces the CRM entirely or sits alongside it in four scenarios:
1. The work has unusual structure no CRM models well.
Recruitment, multi-stage manufacturing, regulated care, complex consulting. The CRM keeps trying to fit the work into "leads, opportunities, deals" and the work refuses to fit. A custom build maps the actual work directly: candidates and placements, jobs and stages, residents and care plans, clients and engagements with their actual structure.
2. The CRM is being used as a customer-facing portal.
Some businesses end up using the CRM to give clients limited access to their own records. That is almost always a poor fit. CRMs are designed for internal sales teams, not external users. A custom client portal that pulls data from the CRM (or replaces the CRM as the source of truth) is the right shape.
3. You are paying CRM vendor lock-in for features you do not use.
Salesforce at full enterprise pricing for a 30-person UK SMB is often £5,000 a month or more, mostly for features the team does not touch. A custom build that captures the 20% of the CRM's features the team actually uses can pay for itself within 12 to 18 months and remove the licence escalator forever.
4. The integration cost is the dominant cost.
If half the operational pain is "the CRM does not talk to the accounts system, the project tool, the support desk, and the email platform", a custom build can absorb the integration logic directly. Often cheaper than buying a CRM with all those connectors and paying per-seat for them.
How to think about cost
A workflow audit costs £2,500 and tells you which of the four scenarios above (if any) apply. The audit fee credits against any build that follows.
A small custom build that replaces a piece of CRM functionality typically costs £3,000 to £8,000. A full custom CRM replacement for a UK SMB usually lands between £8,000 and £25,000, scoped after the audit. Compared to two to four years of Salesforce or HubSpot Enterprise pricing, the maths is usually compelling.
Compared to staying with a CRM that the team is fighting, the maths is even more compelling once you count the operational hours bleeding into it every week.
What to do next
If the parallel spreadsheet exists in your business, the 15-minute triage will tell you in fifteen minutes whether the issue is the CRM, the setup, or the process. The bureau will not pretend a custom build is the answer when it is not.
If you suspect the CRM is the problem but you are not sure, the operational audit is the structured way to confirm. The audit observes the actual workflows, identifies whether the CRM is genuinely the bottleneck, and produces a punch list with fixed-fee proposals if a build is the right answer.
A note on what Orchestrix does not do: the bureau does not configure off-the-shelf CRMs. If the answer is "configure your existing Salesforce properly", the right next step is a Salesforce specialist, not a custom-build consultancy. Half the conversations end with that pointer, and that is fine.