A SaaS CRM at €50 per user per month sounds cheap. Multiplied by 20 salespeople over five years it's €60,000 without counting upsells, add-ons or the real cost of having it badly configured. A well-scoped custom CRM starts to pay off sooner than most founders think, and it gives something SaaS can't: reflecting the business's exact sales process instead of forcing the business to adapt to the vendor's process.
Gartner puts the global CRM market at $91.4 billion in 2024, with Salesforce leading and HubSpot growing double digits year over year (Gartner, 2024). The bulk of that spend is in monthly fees many companies pay without using even 30% of the contracted features.
The five symptoms that SaaS no longer fits
- You're paying for seats no one uses. HubSpot Enterprise runs €3,200/month with 5 seats included (official pricing). If you bought 20 seats and only 8 are active, the real cost per active user triples the list price. It happens more than it seems.
- You have three separate SaaS tools synced via Zapier. Every integration is a failure point. A custom CRM builds into its code the relationships you resolve today with fragile webhooks and Zapier rows that break every month.
- The team enters duplicated data because the CRM doesn't match your process. SaaS forces the operation to adapt to the vendor's model. A custom CRM does the opposite and removes the work of "translating" every movement.
- You need to export to Excel for weekly reports. Generic SaaS reports don't give the KPIs leadership asks for. When ops spends two hours a day exporting and cleaning in Excel, the signal is clear.
- Switching vendors is impossible. If five years of commercial data live in a proprietary model, migration is so expensive that you renegotiate down before switching. That's lock-in — and only a custom CRM avoids it.
When migrating does NOT pay off
If you have fewer than 10 users and standard commercial processes, HubSpot, Pipedrive or Zoho are still the right choice. A well-built custom CRM starts at around €15,000 in initial development plus annual maintenance; with 5 people using it, the break-even takes over two years and the risk doesn't pay off.
It also doesn't pay if the sales team rotates a lot: the learning curve for a custom CRM is steeper than a SaaS with public docs, YouTube courses and an active community.
If five years of your data live in a proprietary model, migration is so expensive that you renegotiate the price down before switching.
How to estimate the real break-even
Add the annual SaaS bill (seats + add-ons + extra integrations) + monthly system-admin hours + cost of fragile integrations + hours of Excel exports. If that number exceeds €25,000–30,000 annually, a €30,000–45,000 custom build with 15% annual maintenance pays off in two years and from year three on it's pure savings.
A real example: MARTLONGENVITY, an integrative medicine clinic that also works in urology and dermatology, replaced a scattered stack of commercial CRM plus one-off integrations plus manual spreadsheet control with a custom CRM integrated with the clinical software. The system reflects the clinic's real patient flow instead of forcing the team to adapt to the vendor's model.
What to keep from SaaS when you migrate
Not everything should be replaced. Outbound tools (Apollo, Lemlist), data enrichment (Clearbit) and e-signature are still better as SaaS than as custom builds. A custom CRM is built for your business's operation, not to replace the whole stack. The useful rule: if the feature is not a differentiator for your business, keep it in SaaS. If it is, build it.
Risks and how to mitigate them
The three big ones: over-specification in the design phase (wanting to cram everything into v1), dependence on a single developer and lack of documentation. All three are mitigated the same way: ship small versions every 2–3 weeks, contract maintenance with the same team that builds it and require an operations manual from the first commit. That's what at udae we call "client ownership from the first commit."
