The Question Every Growing SMB Faces
You've outgrown your spreadsheets. Your team is spending hours on manual processes that should be automated. You need software.
The first question is: do you buy an existing solution or build something custom?
Most SMBs default to buying - it's faster, cheaper upfront, and feels less risky. But "off-the-shelf" is not always the right answer, and the decision deserves more rigour than most businesses apply to it.
The Real Cost of Off-the-Shelf Software
The sticker price of SaaS software is rarely the total cost. Before comparing off-the-shelf to custom, calculate the Total Cost of Ownership (TCO) over 3 years:
Direct Costs
- •Subscription fees: Monthly/annual licence × 3 years
- •Implementation: Setup, configuration, data migration
- •Training: Initial training + ongoing for new staff
- •Customisation: Most SaaS tools charge for custom fields, workflows, or integrations
- •Integration: Connecting to your other systems (often requires middleware or custom development)
- •Support: Premium support tiers for faster response times
Hidden Costs
- •Process adaptation: The cost of changing your processes to fit the software (not the other way around)
- •Workarounds: Time spent on manual workarounds for features the software doesn't support
- •Data lock-in: Cost of migrating away if you switch vendors (often underestimated)
- •Price increases: SaaS vendors typically increase prices 10–20% annually
Example: CRM for a 50-person sales team
| Cost Component | Year 1 | Year 2 | Year 3 | Total |
|---|---|---|---|---|
| Subscription (₹2,000/user/month) | ₹12 lakh | ₹13.2 lakh | ₹14.5 lakh | ₹39.7 lakh |
| Implementation | ₹2 lakh | - | - | ₹2 lakh |
| Training | ₹1 lakh | ₹50,000 | ₹50,000 | ₹2 lakh |
| Customisation | ₹1.5 lakh | ₹75,000 | ₹75,000 | ₹3 lakh |
| **Total** | **₹16.5 lakh** | **₹14.45 lakh** | **₹15.75 lakh** | **₹46.7 lakh** |
The Real Cost of Custom Software
Custom software has a different cost profile: higher upfront, lower ongoing.
Direct Costs
- •Development: One-time cost to build the software
- •Infrastructure: Hosting, database, CDN (typically ₹20,000–₹80,000/year for SMB scale)
- •Maintenance: Bug fixes, security updates, minor enhancements (typically 15–20% of development cost per year)
- •Major enhancements: New features as your business evolves
Hidden Costs
- •Time to build: 2–6 months before you have working software
- •Specification effort: You need to clearly define what you want
- •Vendor risk: If your development partner disappears, you need to find another
- •Technical debt: Poorly built software becomes expensive to maintain
Example: Custom CRM for the same 50-person team
| Cost Component | Year 1 | Year 2 | Year 3 | Total |
|---|---|---|---|---|
| Development | ₹15 lakh | - | - | ₹15 lakh |
| Infrastructure | ₹60,000 | ₹60,000 | ₹60,000 | ₹1.8 lakh |
| Maintenance | ₹2.25 lakh | ₹2.25 lakh | ₹2.25 lakh | ₹6.75 lakh |
| Enhancements | ₹1 lakh | ₹1.5 lakh | ₹2 lakh | ₹4.5 lakh |
| **Total** | **₹18.85 lakh** | **₹4.35 lakh** | **₹4.85 lakh** | **₹28.05 lakh** |
3-year saving vs. off-the-shelf: ₹18.65 lakh
The break-even point is typically 18–24 months for most custom vs. SaaS comparisons.
Beyond Cost: The Strategic Factors
Cost is important but not the only factor. These strategic considerations often matter more:
1. Competitive Differentiation
If your process is a source of competitive advantage, you should own the software that runs it.
Example: A logistics company has developed a unique route optimisation algorithm that reduces delivery costs by 25% compared to competitors. This algorithm is their competitive moat. Building it into a custom system protects it. Using a generic SaaS routing tool means competitors can access the same capabilities.
When this matters: Your process is genuinely differentiated and the software embodies that differentiation.
When it doesn't: Your process is standard (payroll, basic accounting, email) - no competitive advantage in custom software here.
2. Process Fit
Off-the-shelf software is built for the average customer. If your process is significantly different from the average, you'll spend more time and money adapting to the software than the software saves you.
Signs of poor process fit:
- •You need > 20% of the software's features
- •You're using the software in ways it wasn't designed for
- •You have significant manual workarounds
- •Your team complains that "the software doesn't work the way we work"
3. Integration Requirements
If the software needs to integrate deeply with multiple existing systems, custom development often produces cleaner integrations than off-the-shelf tools connected via middleware.
Example: A manufacturer needs their quality system to integrate with their ERP (SAP), their supplier portal, their customer portal, and their IoT sensors on the production line. A custom system can be designed with these integrations as first-class features. An off-the-shelf QMS connected via Zapier will be fragile and limited.
4. Data Ownership and Privacy
With custom software, your data lives in your infrastructure. With SaaS, it lives in the vendor's infrastructure.
When this matters:
- •Regulated industries (healthcare, finance, defence)
- •Sensitive business data (pricing, customer lists, formulas)
- •Data localisation requirements
The Decision Framework
Use this framework to make the build-vs-buy decision:
Step 1: Is this a commodity process?
Commodity processes (payroll, basic accounting, email, standard HR): Buy. These are well-solved problems with mature SaaS solutions. Building custom is wasteful.
Differentiated processes (your unique operational workflow, proprietary algorithm, industry-specific process): Consider custom.
Step 2: Does any off-the-shelf solution fit > 80% of your requirements?
If yes: Buy (with customisation for the remaining 20%).
If no: Custom is worth evaluating.
Step 3: What's the 3-year TCO comparison?
Calculate as shown above. If custom is cheaper over 3 years AND fits your requirements better, custom wins.
Step 4: Do you have the capacity to manage a custom development project?
Custom software requires active involvement from your team: requirements definition, testing, feedback, and ongoing management. If you don't have the bandwidth, a good SaaS solution with some compromises is better than a custom project that fails due to poor management.
The Hybrid Approach
The best answer is often neither pure custom nor pure off-the-shelf:
Buy the commodity, build the differentiator:
- •Use standard SaaS for payroll, email, basic accounting
- •Build custom for your unique operational workflow
- •Integrate them via APIs
Customise an open-source foundation:
- •Start with an open-source platform (Odoo, ERPNext, etc.)
- •Customise for your specific requirements
- •Lower development cost than greenfield, more flexibility than SaaS
Start with SaaS, migrate to custom when you've validated the process:
- •Use SaaS to validate that the process works and understand your requirements
- •Build custom once you know exactly what you need
- •Avoid building custom for a process you haven't validated
Red Flags in the Build-vs-Buy Decision
🚩 "We'll build it because we want full control" - control is not a business outcome; cost, speed, and quality are
🚩 "The SaaS is too expensive" - calculate 3-year TCO before concluding this
🚩 "We'll build it in-house" - unless you have a strong engineering team, this usually ends badly
🚩 "We need it in 4 weeks" - custom software cannot be built in 4 weeks; buy if you need it fast
🚩 "We'll figure out the requirements as we go" - custom development without clear requirements is a budget black hole