AEVOMIND INSIGHTS — SALES & CRM
Bespoke CRM vs off-the-shelf: when custom wins (and when it doesn't)
For most small businesses, an off-the-shelf CRM is the right answer, and a good agency should say so first. Custom wins in a narrower set of cases: when your process does not have the shape a sales pipeline assumes, when the CRM has to share one record with the rest of the business, or when the workarounds cost more than the fit. That last one you can measure in a week.
The short answer
Buy first. An established off-the-shelf CRM is the right answer for most small and mid-sized businesses, and a good software agency should tell you that before it tells you anything else. It is quicker to start, someone else maintains it, and the standard sales pipeline — lead, opportunity, proposal, won — is exactly what these products are built around.
Custom wins in a narrower set of cases, and they share one feature: the cost of working around the product has grown larger than the cost of making something that fits. That cost is measurable, which means this does not have to be a matter of opinion.
What off-the-shelf does well
- Speed to start. You can be entering contacts this afternoon.
- Maturity. Years of other people's feedback have shaped the screens, the mobile app and the email integration.
- Someone else maintains it. Security patches, hosting, backups and new features are not your problem.
- An ecosystem. Connectors to common email, calendar, accounting and marketing tools already exist.
If your sales process looks like the one the product assumes, none of the arguments for building will outweigh these. Configure it properly, train people on it, and stop there.
Where the fit starts to go wrong
These are the signs that a CRM is being worked around rather than used:
- A spreadsheet runs alongside it. The CRM holds the contact; the spreadsheet holds the thing the business actually tracks.
- The same customer is typed in twice. Once in the CRM, again in finance, operations or a job system, and the two drift apart.
- Custom fields have multiplied. Every workaround added a field, and now nobody is sure which ones are current.
- The process is not a pipeline. A college applicant who becomes a student, a lettings applicant who becomes a tenant, a survey that becomes a quote that becomes an installation: these are chains of stages with rules and documents, not a sales funnel.
- Reports are built outside it. The numbers the directors actually use are exported and reassembled every month.
One of these on its own is normal. Several together usually mean the CRM is fine and the problem is that it sits alone, disconnected from the rest of the business. We describe the same pattern in lettings agencies, colleges and installers.
The workaround test
Instead of arguing about features, measure the workarounds. Ask each person who uses the CRM how many minutes a day they spend around it rather than in it: copying into spreadsheets, re-typing into other systems, fixing records, building reports by hand.
Then convert it to staff time: minutes per day × people ÷ 60 for daily hours; × 5 for a week; × 4.33 for a month; ÷ 162.5, the hours in one UK full-time month (37.5 × 52 ÷ 12).
Two worked examples with illustrative inputs:
- Team A: 12 people × 15 minutes a day = 3 hours a day, 15 hours a week, about 65 hours a month — 0.4 of a full-time person.
- Team B: 4 people × 10 minutes a day = about 40 minutes a day, 3.3 hours a week, about 14 hours a month — under a tenth of a person.
Team B should not build anything. Team A has a real problem — but even Team A should check whether connecting its existing systems removes most of the 15 minutes before commissioning a new one.
The inputs are illustrative. Yours will differ, and the only numbers that matter are the ones you measure. The same method underpins how we estimate automation savings generally.
When custom genuinely wins
In our view, a bespoke CRM earns its place when at least one of these is true, and the workaround test says the time is real:
- Your process is the product. The way you move a customer from first contact to delivery is part of what makes you different, and bending it to a generic pipeline loses something.
- It has to share one record with the rest of the business. Sales, finance, operations and compliance need the same customer, the same job and the same history, rather than four copies kept loosely in step.
- The chain has rules. Stages that cannot be skipped, documents that must exist before the next step, deadlines that carry legal weight.
- Most of your users are occasional. People who need one screen, one action or one report, and for whom a full CRM seat is the wrong tool.
The middle path: configure, connect, then build
Between "buy" and "build" sits the option most businesses skip. Keep the CRM, configure it properly, and connect it to the systems it currently shares data with by hand. An integration hub and a shared data layer can remove the re-typing without replacing anything, and workflow automation can carry work across the seams.
Build only the part that is genuinely yours, and keep standard tools for the parts that are not. When a build is right, it should be a CRM built into the same platform as everything it has to talk to, with the data moved across properly through a planned migration rather than a weekend of copying and pasting.
A sensible order of operations
If you are unsure, this sequence costs little and rules things out in the right order:
- Measure for one ordinary week. Minutes per person per day spent around the CRM, and where they go.
- Configure what you have. Remove dead fields, fix the stages to match how you actually work, and train people on the result. Re-measure.
- Connect the worst seam. Usually the one between the CRM and finance or operations. Re-measure again.
- Only then decide on a build, and only for whatever the first three steps could not fix.
Most businesses stop at step two or three, and that is a good outcome. The ones that reach step four arrive with measured requirements instead of a wish list, which is the single best predictor of a build that works.
What to ask before either decision
- Can we export all of our data, in a usable format, whenever we like?
- When our process changes, who can change the system, and how long does it take?
- How will it connect to finance and operations — built in, a connector, or someone re-typing?
- For a build: how is it documented and tested, where is it hosted, and what happens to support if we part ways?
- For a product: which of our workarounds does it actually remove, shown on our data rather than a demo?
If the answers are vague, the decision is not ready, whichever way it is leaning.
Frequently asked questions
Is a bespoke CRM better than an off-the-shelf one?
Not in general, and anyone who says otherwise is selling something. An established off-the-shelf CRM is quicker to start, maintained by someone else and good at the standard sales pipeline. A bespoke CRM is better only when your process does not have the shape those products assume, or when the CRM needs to share one record with the rest of the business. For most small businesses, off-the-shelf is the right first answer.
When does a custom CRM make sense for a small business?
When the workarounds cost more than the fit. The usual signs are a spreadsheet running alongside the CRM, the same customer details typed into finance or operations as well, custom fields bolted on until nobody trusts the data, and reports that are assembled outside the system. If those describe you, measure the time they take before deciding anything.
How do I work out whether our CRM workarounds justify a build?
Ask each person who uses it how many minutes a day they spend working around it: copying into spreadsheets, re-typing into other systems, fixing records. Multiply by the number of people and by five for a week, multiply the weekly hours by 4.33 for a month, then divide by 162.5, the hours in one UK full-time month. Twelve people losing 15 minutes a day is 0.4 of a full-time person. Four people losing 10 minutes is under a tenth.
Can we replace spreadsheets without building custom software?
Often, yes. Many spreadsheets exist because a system was never configured properly, or because two systems do not talk to each other. Configuring the tool you already pay for, or connecting it to the system the spreadsheet is bridging, removes the spreadsheet without a build. Building makes sense for the part that is genuinely specific to how you work.
What are the risks of a bespoke CRM?
You own it, which cuts both ways. Someone has to maintain it, host it, secure it and change it as the business changes, and if only one developer understands it you have a key-person risk. Ask any builder how it will be documented and tested, where it will be hosted, whether you can export all of your data in a usable form, and what happens to support if you part ways. If those answers are vague, the build is not ready.