Choosing a custom ERP or CRM versus a standard SaaS is not “done properly” versus “homemade”. It is how closely the software must match the way you work, and how willing you are to match the software.

SaaS wins on time-to-start, updates and a predictable monthly cost. Custom wins when processes, data or integrations will not fit a configuration screen, and when bending the company to the standard product costs more than the code.

Three options, not two

Horizontal SaaS (well-known CRM or ERP, many industries). You switch it on, train people, pay per user. It works if you sell and administer like thousands of other companies. It stalls when you need statuses, jobs, warehouse or price lists the product treats as permanent exceptions.

Vertical SaaS (industry-specific). Closer to the trade, less flexible outside the vendor’s perimeter. Consider it if the sector is truly covered and APIs exist. The risk is lock-in: data leaves badly, roadmaps are not yours.

Custom (or custom on top of a core). You build what is specific — job costing, a configurator, a shop-floor link, a supplier portal — and you use SaaS or standard modules for commodity (email, basic accounting, if that is enough). This is not “rewriting SAP”. It is refusing to twist the company around a module people hate.

Questions that matter more than feature lists

  • Does the process that differentiates you live in the CRM/ERP, or in how you produce and deliver? If it lives in the process, a SaaS that ignores it makes you work in parallel (Excel, WhatsApp, “the real system”).
  • How many systems must stay connected? Two fragile integrations on a closed SaaS can cost more than an application that speaks REST on purpose.
  • How many users would you pay for over years, including people who log in twice a week? Compounded seats are the real SaaS price.
  • Can you export everything in a usable format? A vague answer means you are buying a rental, not a tool.
  • Do you have someone internal who owns the process, or must the software also “create order” by itself? No product, custom or not, replaces unwritten rules.

When custom is too much

If you only need to track leads and activity, a serious SaaS CRM is enough. If invoicing is ordinary and your accountant already lives in their system, do not rebuild accounting. If the team is small and the process changes every month, custom will eat the budget in rethinks: a simple tool plus discipline is better.

Custom makes sense with a perimeter that can hold for 18–24 months, a process owner, and measurable pain (hours, errors, inability to integrate).

A hybrid path that works

Many SMEs do not “throw away SaaS”. They keep the CRM for the pipeline and build the part the SaaS will not do: a B2B portal, a configurator, ERP sync, a production dashboard. Same rule as custom software in general: pay for specificity, do not reinvent email.

An honest comparison lines up: three-year licences, integrations, training, the parallel work that remains, and exit cost. Without that line you choose by habit (“everyone uses it”) or pride (“we’ll build it”).

DPH designs custom ERP and CRM relative to standard SaaS when the process requires it, and integrates what you keep. To see which side you are on, request a consultation: processes, current systems, and what you still do outside the tool.