Every business that outgrows a spreadsheet asks this, and the answer is usually the opposite of what the person asking expects. Most companies who ask whether to build a CRM should not, and the small number who should are often the ones who have not considered it.
Here is the arithmetic, honestly, including the cases where buying is obviously right.
What a custom CRM costs
Published 2026 ranges put custom CRM development between USD 45,000 and 220,000, and the tiers are fairly consistent across sources:
- USD 45,000 to 80,000 for a first version with contacts, pipelines and activity tracking, typically 12 to 16 weeks.
- USD 80,000 to 150,000 once you add email automation, lead scoring and forms.
- USD 150,000 to 220,000 for a multi-tenant platform with white-labelling and single sign-on.
Against that, HubSpot or Salesforce is a monthly per-seat cost that starts small and grows with your headcount.
The threshold where building starts to win
The consistent finding across the market is that custom becomes defensible at roughly twenty or more seats with a three year horizon, where it typically reaches break-even somewhere between eighteen and thirty months.
Below that it rarely works. For a fifteen person team, a focused build plus annual maintenance almost never beats a well-configured HubSpot instance. The licence looks expensive monthly and is still cheaper than owning software.
The other common trigger is a subscription past roughly USD 2,000 per month, at which point the annual licence starts to approach a meaningful share of a build.
The reasons that are actually good
Cost alone is a weak reason to build. These are stronger:
- Your process genuinely does not fit. Not "we would prefer it differently" but a workflow the platform cannot express without fighting it. This is real in logistics, in construction and anywhere with unusual approval chains.
- The CRM is your product, or part of what you sell to your own customers. Then it is not overhead, it is the thing.
- Data residency or regulatory constraints that a shared platform cannot satisfy.
- Vendor lock-in has become a business risk, usually because pricing changed and you had no leverage.
The reasons that feel good and are not
"We only use ten per cent of Salesforce" is the most common, and it is misleading. You will also use a fraction of what you build, and you will pay to maintain all of it. Unused features on a subscription cost you nothing extra; unused features you own cost you forever.
"We want it to work exactly our way" is worth interrogating too. Sometimes the existing process is habit rather than advantage, and configuring a platform forces a useful conversation about which parts are actually load-bearing.
The cost people forget
A bought CRM is somebody else's responsibility. Security patches, uptime, browser compatibility, compliance updates, integrations with tools you have not adopted yet: all of it is theirs.
Own the software and all of that becomes yours, permanently. Budget maintenance as a real annual line, not an afterthought, and assume you will want changes every year because your business will keep moving.
Ask any vendor for a three year total rather than a build price. One who cannot produce a maintenance figure has not thought about your second year.
A middle path worth considering
The choice is rarely binary. Keep the platform for what platforms are good at, contact management, email, reporting, and build only the piece that genuinely does not fit, connected by API.
You get the workflow you need without owning a whole CRM, and the custom part stays small enough to maintain. For most businesses that ask us this question, this is the recommendation, and it is a smaller engagement than a full build.
How to decide in one sitting
Count your seats. Take your current annual licence cost and multiply by three. Write down the specific workflows the platform cannot handle, being strict about the difference between cannot and prefers not to.
If that list is short and your three year licence is well under a build, configure what you have. If the list is long, load-bearing, and you are above twenty seats, a build is defensible. If you are in between, build only the unfitting piece and keep the rest.