How much does a custom CRM cost? A build-versus-buy read

The question sounds like it wants a figure, and the honest first move is to refuse to give one before the scope exists. But there is a real answer to a better question: not what a custom CRM costs to build, but what it costs to own, because ownership is where the money actually goes and where the build-versus-buy decision is really made.

Start with the thing you are comparing against, because its price is public and yours is not. That asymmetry is the single biggest reason custom CRM decisions go wrong: one option quotes a small monthly number and the other quotes a large one-time number, and they are compared as if they measured the same thing.

What off-the-shelf actually costs

The per-seat prices are low and they are published. HubSpot lists a free tier for up to two users and 1,000 contacts. Salesforce's Sales Cloud pricing runs from a $25 per user per month Starter up to $175 for Enterprise, billed annually. For a five-person team those are numbers a business barely notices.

So why would anyone build? Not to save the seat cost, which custom never does. The reasons that hold up are narrow: a workflow the tools genuinely cannot model, data that cannot leave your infrastructure for a real regulatory reason rather than a nervous one, or a per-seat price that becomes painful at a headcount you can actually forecast reaching. If none of those is true, the rest of this note is an argument for not building.

The build number, and why it is the small one

A genuinely custom CRM, built to a real specification rather than a wish list, is a multi-month engagement. The build is where clients focus because it is the number on the quote, and it is the least important number in the decision, because software that is used is software that changes, and the changes never stop arriving.

The way to keep the build number honest is to keep the first version small, which is the whole argument in scoping a project before anyone writes code. A CRM that tracks contacts, deals and a pipeline is a defined, buildable thing. A CRM that also does email, invoicing, support tickets and reporting is four more products wearing one name, and each one already exists off the shelf.

The costs that arrive after launch

This is the part that decides the real total, and it is the part left off most quotes. A custom CRM has running costs that a subscription hides inside its price:

• Hosting and the database, monthly, for as long as the system lives.
• Maintenance: dependency updates, security patches and the fixes for problems that only appear under real use.
• Changes: every new field, stage or report is a small project, because there is no settings screen a vendor already built.
• The bus factor: if one person understands it and leaves, the next person's first job is to understand it before they can change anything.

None of these is a reason not to build. They are the reason the five-year total, not the build quote, is the number to compare against the subscription. Reporting in particular tends to be underestimated: teams assume a custom CRM comes with custom dashboards, when in practice those are their own build. The cheaper path is often the one in automating manual reporting without buying a BI platform, layered on whatever holds the data.

A five-year comparison

The table is not an argument for either side. It is an argument for filling it in with your own numbers before deciding, because the answer genuinely flips with headcount and with how unusual your workflow really is. This is the same shape as every build-or-buy call on internal tooling: the tool is cheap, the process it enforces is where the value or the cost lives.

When custom is the wrong answer

For most teams under about twenty people, custom is the wrong answer, and it is wrong for a reason that has nothing to do with engineering quality. It is wrong because the off-the-shelf tools have absorbed thousands of workflows and expose the common ones as configuration, and your workflow, however special it feels, is usually one of them wearing local clothes.

The tell is the sentence "but we do it differently". Sometimes it is true and it is exactly the case where custom pays. Far more often it describes a habit rather than a requirement, and the honest test is whether you would defend the difference to a new hire or quietly drop it. If you would drop it, configure the tool to the common path and keep the money.

When custom earns its keep

The cases where building is right share a feature: the CRM is not a record of the business, it is part of how the business runs. A logistics firm whose pipeline is tied to routing, a clinic whose contacts are patients under data rules that forbid a third-party cloud, a firm whose pricing logic is its actual advantage and cannot sit in someone else's schema. In those, the off-the-shelf tool fights the business every day, and the subscription is cheap for software you are constantly working around.

If that is you, the cost question changes from "how much" to "how do we keep it from becoming a second full-time job", and the answer is a small first version, a plan for the running costs above, and the discipline not to rebuild the four products you could have bought. If it is not you, the best custom CRM decision is the one you did not make.