Off-the-shelf CRM or custom: who adapts to whom

The process test: draw your sales flow on a sheet of paper and tick off what a standard Pipedrive configuration already covers. If it covers more than 80%, the case is closed — go SaaS and solve the remaining 20% with team discipline. The balance tips on atypical processes: sales that include technical design work, public tenders with documentation packs, complex agent commissioning, or sales cycles that depend on production capacity.

Integration is the second tipping point. As long as the CRM only has to talk to email and calendar, SaaS connectors are plenty. Once a quote has to pull live prices and stock from the ERP, and a won order has to trigger a production release automatically, synchronisation through generic connectors becomes fragile and expensive to maintain. Building the CRM directly on top of the ERP data removes that entire class of sync errors — and only here does custom earn its price.

Under 10 people in sales and a classic process — lead, qualification, quote, negotiation, contract — there is no honest argument for a custom CRM. We would build for €15,000 what HubSpot gives you for €500 a month, more polished and with a better mobile app than we would write inside that budget. Any firm proposing otherwise at that size is optimising its invoice, not your sales.

There is one more case where custom is a mistake even at larger scale: the sales process is still changing often. A custom CRM freezes the flow in code — every process change becomes a development order. Stabilise the process on a flexible SaaS first, for a year or two, and build custom only once the flow has settled and the platform's limits hurt you monthly, measurably.