Custom ERP vs generic SaaS: how to decide for a Colombian SMB

What is an ERP and why does it usually fail with SMBs?

An ERP (Enterprise Resource Planning) is a system that unifies business operations: sales, inventory, accounting, invoicing, HR, and everything your company does. The idea is that all data lives in one place and decisions are made with up-to-date information, not with the Excel Pedro emailed yesterday.

The problem is that enterprise ERPs (SAP, Oracle, Dynamics) were designed for companies with 500+ employees, specialized departments, and six-figure USD implementation budgets. When a 15-person SMB tries to implement an ERP at that level, one of three things happens: they abandon the project halfway, they implement it but use 20% of its capacity, or it costs so much to maintain that they end up going back to Excel.

That's why SMB SaaS exists: designed for small teams, simple onboarding, priced per user per month rather than a big upfront license. They're the right choice for most small businesses.

When does a generic SaaS make sense?

In Colombia the most common options for SMBs are:

• Siigo, strong in DIAN electronic invoicing, basic accounting, and payroll. Default for accountants and bookkeepers. Monthly cost per user.
• Alegra, similar to Siigo with a more modern UX. Invoicing, basic inventory, bank integration, reports. Very popular with entrepreneurs and small shops.
• Odoo, modular open-source ERP with a paid version. More powerful than the previous two but requires more configuration. Good when you need serious inventory or CRM.
• World Office, ContaPyme, Helisa, more traditional options in the Colombian market, oriented toward accounting first.

SaaS is the right choice when: (1) your processes are standard (you sell products or standard services), (2) your volume is still growing, (3) you don't have an internal tech team, (4) you need DIAN electronic invoicing without pain, and (5) you can pay between $50K and $500K COP monthly without it hurting.

If you fit that profile, building custom is expensive, slow, and probably unnecessary. Start with a SaaS. You can always migrate later.

When does custom make sense?

There are four situations where a custom ERP clearly beats SaaS:

1. Your operation has processes no SaaS models well. If you sell land with direct financing, rent heavy machinery by the hour, track patients through a medical program with specific protocols, or anything where SaaS forces you to "adapt your operation to the system" rather than the other way around, that's the clearest symptom.

2. You already used a SaaS and outgrew it. Not "I could have used one and decided not to", but "I used Odoo/Siigo/Alegra for three years and hit its limits". At that point you know exactly what you need and what you don't, and that clarity is half the work of building custom.

3. Monthly SaaS cost became a problem. SaaS charges per user per month. If you have 20 people accessing and pay $20 USD each, that's $4,800 USD per year, $24,000 USD over five years. A well-thought-out custom ERP costs half of that upfront and nothing recurring after.

4. You want the code and data to be yours. For critical operations (healthcare, finance, legal), depending on the SaaS to keep existing, not raising prices, or not changing the rules is a real risk. A custom ERP running on your own VPS is literally your system.

How much does each path really cost?

SMBs tend to underestimate the real cost of SaaS and overestimate the cost of a custom ERP. Rough numbers in Colombia:

• SaaS for 5-20 users: $50K - $500K COP monthly = $600K - $6M COP per year = $3M - $30M COP over 5 years. Scales linearly with team size.
• Custom ERP with 2-3 modules: high upfront cost (development), but very low monthly cost (only hosting and optional support). The break-even with SaaS usually lands between 18 and 36 months, depending on team size.
• Opportunity cost: SaaS lets you start tomorrow. A custom ERP takes weeks or months. That time sometimes matters more than any other variable.

A rule of thumb we use: if you'll be on the system for more than 2 years and have more than 15 people using it, a custom ERP pays for itself. Below those thresholds, SaaS is usually more efficient.

What signals indicate you're ready for custom?

Three symptoms we've seen in every SMB that decided to build:

• "We export to Excel to do X". If every day someone exports data from the SaaS to a spreadsheet to do calculations or reports the SaaS can't do, your operation has already outgrown what the SaaS covers.
• "We have another tool for Y that doesn't talk to anything". If your team uses the SaaS for accounting, a different system for accounts receivable, an Excel for collections, and WhatsApp to coordinate, you have silos a custom ERP could unify.
• "The SaaS doesn't let us do Z". If there's a specific flow your business needs and the SaaS doesn't support it (neither via configuration nor plugins), and that flow is core to your operation, it's the clearest signal.

Team Alord (one of our clients) came to us with all three symptoms at once: they exported accounts receivable to Excel because no SaaS modeled "lot with monthly payment plan", they had silos between receivables/finance/machinery, and the delinquency-alert flow simply didn't exist in the SaaS tools they tried. Building custom was the only coherent answer.

Decision framework at a glance

If you want it in concrete questions:

• Just starting? → SaaS
• Is your operation standard? → SaaS
• Team of <15 people? → SaaS, probably
• Exporting to Excel every day? → Think about migrating or building
• Have 3+ information silos? → Custom ERP starts making sense
• Does the SaaS fail to model your operation? → Custom ERP
• Used SaaS 2+ years and outgrew it? → Custom ERP is the natural answer
• Need the data and code to be yours? → Custom ERP, no discussion

Conclusion

The most common mistake we see in SMBs is jumping to a custom ERP before validating that SaaS doesn't work for them. It's expensive, slow, and often unnecessary. Start with a cheap SaaS, use it until you find the friction points, document what you'd want to work differently, and then build, by then you'll know exactly what to build.

The second most common mistake is staying on SaaS when the operation has already outgrown it. Exporting to Excel every day, running three disconnected systems, continuing to pay per user as the team grows, all invisible costs that add up to more than the custom ERP they never built.

The right decision depends on which phase you're in.