Custom ERP vs Off-the-Shelf SaaS for SMBs

The real trade-off: fit, total cost, lock-in

The question "custom ERP or off-the-shelf SaaS" is badly framed when reduced to a comparison of purchase prices. The right trade-off rests on three axes that actually decide the real value of the software over five years: its fit with your business processes, its total cost of ownership, and the degree of lock-in it creates toward a vendor. This is the angle we develop in depth in our complete guide to custom ERP, the pillar page that links this comparison to all our resources on business software.

Fit is the first axis, and the most structuring. An off-the-shelf SaaS is designed for a market's greatest common denominator: it offers configuration options, but you remain bounded by what the vendor planned. If your way of producing, invoicing or coordinating your teams falls outside the standard frame, you have two choices, both costly: twist your process to make it fit the tool, or stack workarounds (Excel exports, double entry, side tools) that recreate exactly the chaos the software was supposed to remove. A custom ERP inverts the logic: it is the software that fits your flows, your calculation rules and your business vocabulary, not the other way round.

The second axis is total cost of ownership, and it is the most misunderstood because it gets confused with the entry ticket. An off-the-shelf SaaS shows a low initial cost, but most often bills per user per month: the tab climbs mechanically with your headcount and time, without ever stopping. A custom ERP requires a concentrated initial investment, EUR 30,000 to 90,000 for a first structuring business scope, then holds at a predictable and low cost, with no recurring licenses to renew. Beyond a certain number of users and years, the two curves cross and custom becomes more profitable.

The third axis, vendor lock-in, is the quietest and the most dangerous. With an off-the-shelf SaaS, your data lives at the provider's, its reversibility depends on their terms, and any evolution of your business assumes the vendor is willing to plan it on their roadmap. If they raise their prices, change strategy or disappear, you suffer. Take a concrete example: for a resort we support in Thailand, no standard hospitality package covered the site's real operations without contortion. Adopting a SaaS would have meant bending the resort's workings to a generic tool and depending on a distant vendor. Custom made it possible to keep control of both the software and the data. These three axes are not independent: a tool that fits your processes poorly is also the one that multiplies workarounds, therefore hidden costs, therefore lock-in. Looking at them together avoids deciding on the sticker price alone.

Point-by-point comparison: custom versus off-the-shelf

Here are the seven criteria that genuinely weigh in the decision. None is absolute: a good off-the-shelf SaaS stays unbeatable on a perfectly standard need, and a poorly scoped custom ERP can drift. The table reflects the common scenario of an SMB whose core business falls outside the generic frame. Across the seven rows, four lean clearly toward custom (business fit, recurring cost, data ownership, evolvability), two toward off-the-shelf (initial cost and time-to-market), and one depends mostly on the quality of execution (integrations).

The most underestimated line is the recurring cost. An off-the-shelf SaaS subscription billed per user per month looks harmless for five people; at forty users over five years, the total often exceeds what a custom foundation amortized over the same period would have cost, without ever making you the owner of anything. The second underestimated line is data ownership: as long as all goes well, you do not think about it; the day you have to migrate or audit, it becomes decisive. That is why we detail these items on our page dedicated to the price of a custom ERP, which prices each variation factor.

On integrations, the table classes custom as fully open, and that is accurate: we connect your accounting, your payroll or a payment tool like Stripe via API, without depending on a catalogue of connectors. An off-the-shelf SaaS offers ready-made connections, handy when they exist, but impassable as soon as the brick you need is missing. Here again, the right criterion is not "which has the most connectors" but "which covers exactly mine".

When off-the-shelf is enough, when custom becomes essential

Choosing means arbitrating by real usage, not by a preference of principle. Here is an honest decision grid, with no commercial bias. Our conviction is simple: you pay for custom only where it creates a real competitive advantage, and you keep off-the-shelf everywhere else.

Off-the-shelf SaaS stays the right choice in these cases

1. Your process is standard and a mature product covers it without contortion: accounting, payroll, e-signature, email.
2. You have a small number of users and a need to start immediately, where time-to-market outweighs a perfect fit.
3. The function is not strategic: if the tool changes, your competitive advantage does not suffer, only the comfort drops a little.

A custom ERP becomes essential in these cases

1. Your processes are your competitive advantage and no generic tool covers them without forcing you to bend your business.
2. You accumulate several poorly connected SaaS tools, with double entry and re-entry, and you pay licenses that climb with your headcount.
3. Data ownership and security are decisive, notably in sensitive sectors where reversibility cannot depend on a third party.

A telling example: a five-person SMB invoicing standard services has no reason to code its own invoicing tool; an off-the-shelf SaaS is perfect and its cost stays anecdotal. Conversely, a resort whose operations fit into no package gains from mastering every flow. The line is therefore not the sector or the size, it is the strategic stake attached to the process concerned.

An intermediate case deserves to be named, because it is very common: the SMB that starts with a pile of SaaS out of budgetary caution, then hits the glass ceiling after two or three years. Each tool works, but they do not talk to each other: teams re-enter the same data into three interfaces, the figures diverge from one tool to another, and the cumulative subscription bill ends up matching what a custom foundation would have cost. It is not a failure of SaaS, it is a poor alignment between the tool and the real maturity of the company. The honest diagnosis consists of anticipating where the organization will stand in three years, not just on the day of the choice.

That is why the most relevant approach is often hybrid. You keep the SaaS tools that do their job very well (accounting, payroll, support) and you develop the one specific business core in custom, linking everything by API to remove double entry. This articulation is among the factors that make the price of an ERP vary, since each connection requires development, but it avoids rebuilding everything just to code what brings a real advantage.

Our decision rule fits in one question: if this tool disappeared tomorrow, would the company lose a competitive advantage, or only a bit of comfort? If the answer is "an advantage", the process deserves a foundation you fully control. If the answer is "a bit of comfort", the fastest SaaS to deploy does the job just fine. This question avoids the two symmetrical mistakes: over-investing in a mundane function, or under-investing in the process that carries the margin.

Our custom stack: Supabase, Stripe, Vercel

When we build a custom ERP, we do it with a stack we stand behind and master end to end, chosen for security, reliability and data ownership. Four bricks structure our approach and answer directly the three trade-off axes.

1. Next.js and TypeScript in strict mode for the application foundation: typing that catches errors before go-live and reduces the cost of maintenance over time.
2. Supabase with RLS and multi-tenancy for the data: strict isolation via row-level security rules, GDPR compliance and hosting you control, unlike a SaaS where your data lives at the vendor's.
3. Stripe and Resend or Brevo for invoicing and notifications: proven bricks connected by API rather than rebuilt, in the hybrid spirit we recommend.
4. Vercel for deployment: automatic go-live at each evolution, preview before publication and instant rollback if needed.

This stack is not a stylistic flourish: it materializes the fundamental difference between custom and off-the-shelf. Supabase and its isolation via RLS answer the data-ownership axis; the absence of a per-user license answers the total-cost axis; transmissible, legible and documented code answers the lock-in axis, because another competent provider can take over the tool without rebuilding everything. It is this continuity we applied on the ERP of a resort, and the same rigor runs through our vertical SaaS products built in-house, such as CoProFlex for condominium syndics or DocAgora in healthcare.

Concrete case: the custom ERP of a resort

To anchor this trade-off in the real world, nothing beats a concrete case. A resort project illustrates exactly the situation where off-the-shelf shows its limits and where custom becomes the right choice: a business whose operations fit into no standard package, field teams and an administration to synchronize, and a need for full control of both the software and the data.

What to remember from this case is the decision sequence. The resort was coordinating its operations via a mosaic of disparate tools and spreadsheets: an assembly of generic bricks, exactly the poorly managed hybrid scenario described above. Adopting an off-the-shelf hospitality SaaS would have imposed bending atypical operations to a frame designed for others, with its share of workarounds and a dependence on a distant vendor. Choosing custom made it possible to centralize steering in a single interface, structured around the teams' real flows and integrated with the existing system. It is the concrete translation of the fit axis: the software adapted to the resort, not the other way round.

This case is not separable from our method. We approach every project through business workshops that map the flows before a single line of code, exactly the approach detailed on our page about the custom ERP development methodology. It is this scoping that avoids the over-engineered mess and guarantees that custom stays focused on what creates an advantage, without re-coding what an off-the-shelf SaaS would already do very well.

In short, reason in total cost of ownership over five years, never in purchase price nor in monthly subscription cost taken in isolation. List the genuinely strategic processes, estimate the number of users and the duration of use, compare the sum of the licenses to the investment of a custom foundation of EUR 15,000 to 150,000, and decide axis by axis. For a standard need and few users, off-the-shelf SaaS is legitimate. For a business core that makes your difference, custom costs less where it counts and makes you the owner of your tool. If you are still hesitating, a free diagnosis makes it possible to price both scenarios on your precise case before any commitment.