Custom software, SaaS, or ERP: how to decide without regret

Short answer: use off-the-shelf SaaS when your process looks like everyone else's. Use an ERP when you need a large set of standardized modules (tax, inventory, finance). Build custom when the way you operate is what sets you apart, or when you already spend more bending the off-the-shelf tool than building would cost. When in doubt, start with the off-the-shelf tool and build only what it cannot do.

A custom software vendor arguing for custom software has an obvious conflict of interest. So we start by admitting what is true: for many companies, off-the-shelf is the right choice. We would rather you know that now than find out after signing.

The three options in one table

Five questions to decide

1. Is your process common, or is it your edge?

If what you do is what most companies your size do (issue invoices, track payables, handle tickets), good, cheap tools already exist. If the way you price, schedule, produce, or deliver is why customers choose you, forcing it into a generic tool can cost the very advantage you have.

2. How many workarounds do you already use?

Classic sign: the team uses the off-the-shelf tool plus a spreadsheet next to it, or types the same data into two screens. When a tool only works with hacks, the cost of living with it shows up in people's hours, not on the invoice.

3. How many people will use it?

Many SaaS products charge per user. With 5 users, off-the-shelf is hard to beat on price. With 80, the math changes. See the numbers just below.

4. Does it need to talk to other systems?

Plenty of SaaS tools and ERPs integrate with banks, invoicing, online stores, and messaging apps. But if your operation depends on connecting five things in a very specific way, custom software can be a cleaner glue than a web of tools linked by fragile automations.

5. How much control do you need over your data and roadmap?

With SaaS, the vendor decides what the product gains or loses next year, and can change price or plans. With custom software, the code and data are yours (as long as the contract says so). That autonomy is worth a lot to some businesses and nothing to others.

The three-year math (example with invented numbers)

SaaS prices vary widely by product. The amounts below are hypothetical and only show the logic; they do not describe a real product or client. Amounts are in Brazilian reais.

Scenario A: 15 people.

• SaaS at R$ 89 per user per month: 15 x R$ 89 = R$ 1,335 a month, R$ 16,020 a year, R$ 48,060 over 3 years.
• Custom: R$ 90,000 to build plus R$ 1,500 a month for hosting and maintenance (R$ 54,000 over 3 years), R$ 144,000 over 3 years.

Result: SaaS is much cheaper. If off-the-shelf does the job, buy it.

Scenario B: 80 people, same scope.

• SaaS: 80 x R$ 89 = R$ 7,120 a month, R$ 85,440 a year, R$ 256,320 over 3 years.
• Custom: the same R$ 144,000, because the cost of building does not grow in step with the number of users.

Result: custom becomes cheaper over three years, and fits the process too.

The lesson: the tipping point depends on user count, how much you adapt the off-the-shelf tool, and the time horizon you use. Run it with your own numbers, including the cost of the hours spent on workarounds. And remember the build cost itself depends on scope, as we explain in how much does custom software cost.

Where each option tends to go wrong

SaaS: you buy from the demo and discover in week two that the report your accountant needs is missing. Or the plan that has the feature costs three times more.

ERP: the rollout becomes a year-long project, the team uses 20% of the modules, and the rest ends up in a spreadsheet on the side. That is not an ERP defect; it is what happens when the company's process and the system's default do not match.

Custom: scope too big from day one, no discovery, nobody on the client side to make decisions, and a vendor who disappears after delivery. All avoidable, all common. That is why we wrote a checklist for choosing a software house.

The middle path that often works

You do not have to pick one option for everything. Many companies combine:

• Off-the-shelf for the common stuff: finance, payroll, email, invoicing.
• Custom for the core of the business: the process only you run your way.
• Integration between the two, so data is not typed twice.

That way you do not pay to build what already exists, and you are not stuck with off-the-shelf where it does not fit.

Signs it is time to look at custom

• The team keeps a "shadow" spreadsheet for what the system cannot do.
• The same data is typed in two or three places.
• Your process changed in a way that gives you an edge and the tool cannot follow.
• Per-user cost grows faster than revenue.
• You hear "the system can't do that" more than once a month.

And signs it is not the time: the process is still being figured out, nobody can describe how the work is done today, or the budget only covers the off-the-shelf tool. In those cases, an inexpensive SaaS teaches you what you really need, and that information is worth gold if you later decide to build.

Frequently asked questions

Is custom software always better than off-the-shelf?

No. It is better when your process is specific, or when user count and adaptations make the off-the-shelf tool expensive. For common processes and small teams, off-the-shelf is often the best choice.

Can I start with SaaS and migrate later?

Yes, and it is often the best path. Just make sure you can export your data in an open format, because migration depends on it.

Does an ERP solve everything?

It handles the standardized parts well (tax, accounting, inventory). What is specific to your business almost always needs expensive customization or a system alongside it.

How long does custom software take?

A lean first version usually takes 2 to 4 months, according to market estimates; the real timeline depends on scope.

Next step

If you are torn between the three and want an honest second opinion (including "buy the off-the-shelf one"), tell us how your process works today. We will say where off-the-shelf fits and where building would make sense.

Talk to us