When a custom CRM is cheaper than a subscription
Per-seat pricing, fields nobody fills in, and a process you bent to fit the tool. A rough model for deciding whether to build instead of buy.
Buying is usually right. Twenty seats of a good CRM is cheap next to building one, and somebody else handles the upgrades. But there is a point where the maths flips, and it is more common than vendors would like.
The three signs
Your team keeps a spreadsheet next to the CRM. Your process has a step the tool cannot express, so people skip it. And your monthly bill grows with headcount while the value per seat does not.
What building actually costs
A focused CRM — pipeline, contacts, activities, roles, reports — is six to ten weeks, not a year. What people underestimate is the second year: updates, small requests, a new integration. Budget for that or the system rots.
Build the ugly 20% first
Start with the part that is specific to you and impossible to buy. Keep the generic parts boring and familiar, because that is what makes the team adopt it.
A simple test
If a custom build pays for itself in under two years of subscription cost, and your process is genuinely unusual, build. Otherwise buy and spend the money on integration instead.
I take on a small number of projects at a time. Tell me what you are building and I will tell you honestly whether I am the right person for it.
Get in touch →