Issue 12EssayNo. 001
HubSpot or a custom CRM? When to build.
Most businesses should buy a CRM. A few should build one. The signs you're in the second group — and the hybrid that often beats both.
Rafael Santos
Partner · Engineering
- Published
- Updated
- Read time
- 8 minutes

In this article
We build custom CRMs for a living, and we tell most people who ask for one not to build it. Platforms like HubSpot are excellent at what most sales and marketing teams need: contacts, pipelines, email, reporting and a huge ecosystem of integrations. Building your own means owning every one of those features, forever.
But some businesses outgrow off-the-shelf CRMs in ways no amount of configuration fixes. When that happens, a custom system can be the cheaper, faster option over a few years. The trick is telling the two situations apart before you've spent a year bending a platform into a shape it resists.
Start with a platform if…
- Your sales process looks broadly like everyone else's: leads, qualification, proposals, deals won and lost.
- Marketing automation — email, forms, landing pages, lead scoring — is a big part of the value.
- Your team is small enough that per-seat pricing stays comfortable for the next few years.
- You'd rather pay a subscription than employ or retain developers to maintain software.
For these teams, the fastest route to value is a well-configured platform, clean data and a team that actually uses it. Our 30-day CRM migration plan is written for exactly that path.
Signs you've outgrown it
These are the patterns we see when clients come to us after a few years on a platform.
Your core record isn't a deal.
If the business revolves around shipments, properties, patients, projects or equipment, each with its own lifecycle and relationships, you'll spend a long time modelling them with custom objects and workarounds.
The CRM is a hub of integrations that fight each other.
When pricing lives in an ERP, scheduling in another tool and the CRM syncs with all of them through a chain of connectors, every change risks breaking something.
You pay for seats people barely use.
Field teams, partners or operations staff who update one record a day can make seat-based pricing expensive, quickly.
The process is your advantage.
If the way you quote, schedule or serve customers is what sets you apart, a generic workflow flattens it.
Data ownership is non-negotiable.
Regulatory, contractual or strategic reasons to keep all customer data in your own infrastructure narrow your options.
One sign on its own rarely justifies a build. Three or more usually means it's worth costing properly.
Separate platform limits from configuration debt, though. Some teams believe they've outgrown HubSpot when what they've really outgrown is years of unmanaged properties, overlapping workflows and duplicate records. A clean-up costs far less than a build, so try it first.
Cost it over three years, not one
Comparing a monthly subscription with a build quote is misleading. We compare total cost over three years, line by line.
| Cost line | Platform | Custom build |
|---|---|---|
| Up-front | Setup, configuration, migration | Discovery, design, build, migration |
| Recurring | Seats, tier upgrades, paid add-ons | Hosting, monitoring, support retainer |
| Integrations | Connector subscriptions and upkeep | Built in and maintained with the app |
| Change | Admin time, within the platform's limits | Developer time, within your budget |
| Exit | Data export and re-implementation | You own the code and the data |
Fill it in with real quotes and your actual headcount plan. Platform costs tend to rise with seats and tiers as you grow; custom costs are front-loaded and then flatter. Where the two lines cross — if they do — is the honest answer.
“Buy the commodity, build the advantage. The trouble starts when a business treats its advantage like a commodity.”
The hybrid that often wins
Many of our CRM projects aren't either-or. Marketing stays in HubSpot, where email, forms and attribution are hard to beat. The operational core — the part unique to the business — becomes a custom application, often built on Laravel, that syncs contacts and deal status with HubSpot through its API. Sales sees one timeline; operations gets a tool shaped around its work.
The hybrid needs clear rules about which system owns which data. Decide that up front — contacts and marketing consent in HubSpot, jobs and pricing in the custom app, for example — and keep each sync one-directional wherever you can.
Treat the sync as part of the product: log every sync, alert on failures and give someone a simple screen to replay anything that didn't go through. Most of the hybrid problems we're called in to fix are silent sync failures nobody noticed for weeks.
If you do build
- Start with the one workflow the platform handles worst, not a full CRM replacement.
- Design the data model with the people who use it daily, then build the screens.
- Plan for reporting, permissions and audit logs from the start; they're hard to retrofit.
- Budget for maintenance every year. Software you own is software you look after.
We usually build these on Laravel: it's mature, well documented and easy to hire for, with first-party tooling for queues, authentication and admin panels. Any mainstream framework your team can maintain is a reasonable choice, though. The stack matters far less than the data model.
A custom CRM done well becomes an asset the business runs on. Done for the wrong reasons, it's an expensive copy of a subscription. If you're weighing the two, our CRM development team will cost both honestly — including when the answer is to stay where you are.
Key takeaways
If you only remember three things
Most teams should buy. Build when your core record, process or data rules are genuinely unusual.
Compare three-year total cost, including integrations, change and exit.
Consider the hybrid: HubSpot for marketing, a custom app for the operational core.

Written by
Rafael Santos
Partner · Engineering · 15 years shipping software
Rafael leads engineering across our web, app and CRM projects. He has shipped products on most of the stacks clients ask us about, and writes here about the technical decisions that age well — and the ones that don't.


