Case study: the custom CRM that replaced HubSpot for a 12-person tech company

Retool's 2026 Build vs. Buy report says 35% of enterprises have already replaced at least one SaaS tool with custom-built software, and CRM is near the top of the list. (Newsweek) That number tracks with what we're seeing from smaller companies too. Here's a recent one.

The setup

A 12-person B2B software company came to us paying around $34,000/year for HubSpot across sales, marketing, and service hubs. They'd been on it for four years. The founder told us, plainly, that nobody on the sales team actually opened it unless they had to. Deals lived in a shared Notion page. Forecasts lived in a spreadsheet. HubSpot was where data went to be ignored before the quarterly board meeting.

This is the part nobody puts in a case study: most CRM problems aren't CRM problems. They're workflow problems wearing a CRM costume.

What they actually needed

We spent two afternoons watching how the team worked. The real list was short:

- A pipeline view three people could update without a training video - Automatic enrichment from their product's signup data (the CRM didn't know who was already a trial user) - A weekly digest the founder could read on Sunday night - Notes that didn't require seven clicks to find later

They did not need marketing automation, ticketing, a chatbot builder, or 90% of what they were paying for.

What we built

A single internal app. Postgres database, a thin React frontend, and a handful of background jobs. Auth through their existing Google Workspace. We pulled trial signups from their product's API, deals from the team's Notion (until we deprecated it three weeks later), and Stripe data for closed-won amounts. The weekly digest is a Sunday cron job that drops a markdown email in the founder's inbox.

Time from kickoff to the team using it daily: 19 days. Cost to build: less than four months of the HubSpot bill. Ongoing cost on our managed infrastructure: about $90/month in compute and storage.

What changed

Three things worth naming:

1. The team actually uses it. Adoption isn't a slide we present; it's the fact that the pipeline updated 41 times in the first week without anyone being asked. 2. Forecasts stopped being fiction. Because the data flows in automatically, the Sunday digest reflects reality instead of whatever someone remembered to type on Friday. 3. The founder stopped thinking about the CRM. Which is the point. Good automation is invisible.

The honest caveat

This worked because the company is small and their process is stable. If they were running 50 reps across four regions with complex commission rules, we'd have a different conversation, and the answer might genuinely be Salesforce. Custom isn't always right. It was right here.

If you're staring at a renewal quote and wondering whether the tool is worth what it costs you, that's the conversation we like having. No pitch, just diagnosis.