Custom vs. off-the-shelf software: when each one wins

It is the first decision in almost any software project and the one that can save the most money — or burn it: do I buy a system that already exists or do I build my own? After years implementing both paths for universities, clubs and companies in Peru, this is the honest answer: it comes down to a single question — whether the process you want to systematize is a commodity or your differentiator. Everything else follows from there.

First, the terms, clearly

Off-the-shelf software is a product that already exists and is used as-is —an ERP, a CRM, a payroll system— for a license that is almost always recurring and per user. Custom software is designed for your specific operation: the heavy investment is up front and the result, source code included, is yours.

Off-the-shelf software (also “stock”, COTS or SaaS) is a product that already exists and is used as-is: an ERP like SAP Business One, a CRM like HubSpot, a payroll system like Buk. You pay a license — almost always recurring and per user — and in return you get a mature product, maintained by its vendor, that solves a standard process.

Custom software is a system designed and built for your specific operation: your rules, your workflows, your integrations. The heavy investment is up front (the development), the recurring cost is maintenance, and the result — source code included — is yours.

The real cost: don’t compare the entry price

The serious comparison isn’t the monthly fee against the development budget, but the total cost of ownership over 3-5 years. With off-the-shelf, the per-user license multiplies and never ends; with custom, maintenance runs around 15-20% a year and doesn’t grow with your team.

The most common mistake is comparing the off-the-shelf monthly fee against the custom development budget and deciding on that. The serious comparison is the total cost of ownership (TCO) over 3–5 years:

With off-the-shelf, the per-user license multiplies: 50 users × monthly rate × 60 months, plus the “premium” modules you discover you need, plus the cost of adapting your process to the product (training, workarounds, parallel spreadsheets). The entry price is low; the cost grows with your team and never ends.

With custom, the curve is the opposite: a heavy up-front investment, then maintenance typically at 15–20% a year of the development cost. You don’t pay per user: if your operation grows from 50 to 500 people, the software costs the same. The break-even point usually arrives between the second and fourth year — sooner if you have many users, later if you have few.

When off-the-shelf is the right call

Off-the-shelf software makes sense when the process is standard in your industry —accounting, payroll, e-invoicing—, you have few users and you need to be operating in weeks. If the process doesn’t set you apart from your competition and a mature product already solves it, building custom is overspending.

Let’s be clear: for commodity processes, off-the-shelf wins almost every time. Accounting, e-invoicing, payroll, email, digital signature: there your process isn’t special (nor should it be — SUNAT regulates it, not your creativity), and competing against a product with thousands of clients and years of maturity is throwing money away.

When custom is the right call

When the process is your differentiator or your critical operation and no product covers it without bending it out of shape. The four signals: deep integrations between systems, business rules that don’t fit in a configuration form, particular scale or concurrency, and the need to own the code and the data.

Custom is justified when the process is your differentiator or your critical operation and no product covers it without bending it out of shape. The four most reliable signals we see in practice:

1. Deep integrations. When the value lies in connecting systems — the academic system with the cashier, enrollment with the ERP, identity with RENIEC — off-the-shelf falls short: those connections specific to your operation don’t exist in any catalog. In the cashier system we built for a national university, the key piece wasn’t collecting payments: it was that every other system validated each receipt in real time against our own APIs, with identity verification against RENIEC. That doesn’t come off the shelf.

2. Your own business rules. If your rules fit in a configuration form, use off-the-shelf. If you need “only the manager can void, only on the same day, and the refund document is generated automatically”, you’re describing your own code.

3. Particular scale or concurrency. A club with more than 50,000 members asked us for an enrollment system that would survive peak day, when thousands log in at once. The solution was a virtual waiting room with numbered turns and wait times — a requirement so specific to their operation that no generic product shipped with it.

4. Ownership of the code and the data. With off-the-shelf, your operation lives on a third party’s platform: if they raise the price, discontinue the product or restrict the API, you inherit it. With custom, the system is an asset of yours — it amortizes, it can be audited, and it evolves at your pace.

The middle ground almost nobody tells you about: hybrid

The real decision is almost never all-or-nothing. The pattern that works best in mid-sized and large operations is off-the-shelf for the commodity, custom for the core, and the two integrated: accounting in an ERP and the differentiating process in your own system that sends it every transaction.

The real decision is almost never all-or-nothing. The pattern that works best in mid-sized and large operations is off-the-shelf for the commodity, custom for the core, integrated. That same enrollment club is the perfect example: invoicing runs on SAP Business One (off-the-shelf — accounting is commodity), but enrollment, its differentiating operation, is a custom system that sends every transaction to the ERP automatically. Nobody keys anything twice. Each piece does what it does best.

The two classic mistakes

Over-customizing an off-the-shelf product —you pay licenses and perpetual consulting, and in the end nothing is yours; if the customization exceeds ~30% of the product, it was a custom project in disguise— and building custom what is a commodity, like your own accounting or your own payroll.

Over-customizing an off-the-shelf product. It’s the worst of both worlds: you pay recurring licenses and perpetual consulting, the product fights against every vendor update, and in the end nothing is yours. If the customization exceeds ~30% of the product’s behavior, it was a custom project in disguise.

Building custom what is a commodity. The mirror image: developing your own accounting, your own email or your own payroll is reinventing — on your budget — something the market has already perfected for a fraction of the cost. A good custom software vendor tells you where hiring them isn’t worth it.

A checklist to decide in 10 minutes

Five questions: does this process set me apart or is it the same across my whole industry?, is there a product that covers 80% without bending it?, do I need deep integrations with RENIEC, SUNAT or my ERP?, how much does the per-user license add up to over three years?, and what happens if the vendor raises the price or shuts down?

In short

Buy the commodity, build your differentiator and integrate them. Off-the-shelf is unbeatable for standard processes with few users and urgency; custom wins when the process is your advantage, the integrations are deep or the scale is yours alone.

Buy the commodity, build your differentiator, and integrate them. Off-the-shelf is unbeatable for standard processes with few users and urgency; custom wins when the process is your advantage, the integrations are deep or the scale is yours and nobody else’s. And if a vendor always pushes you toward the same side — either of the two — be wary: the right answer depends on your operation, not on what they sell.