Every growing business reaches a point where a spreadsheet, a messaging group and a generic SaaS tool no longer match how the work actually happens. The question is not whether to change the tool. It is whether to adapt a product that already exists or commission software built for your process.
Build versus buy is rarely a binary choice. Most sensible decisions sit between full custom development and pure off-the-shelf purchase. This guide sets out the criteria, the trade-offs and the decision rules we use when scoping work for SMEs in Bangladesh.
Start with the process, not the technology
Write down the process that hurts. Include who does what, where the handoffs fail, how often exceptions occur and what a mistake costs. If the process is standard and the pain comes from poor adoption, a new build will not fix it. If the process is genuinely different and the workaround is where customers feel the delay, custom software may be justified.
A useful test: can you describe the process in a way a new employee could follow in one page? If not, fix the process first. Software automates whatever you document, including the confusion.
The criteria that matter
1. Is this process a source of advantage?
Buy for commodity processes: accounting, payroll, email, standard CRM, document storage. Build where the process is part of your value proposition, such as a specialised production workflow, a unique pricing model or a service promise competitors cannot copy easily.
2. What is the three-year total cost?
Compare on the same basis. For buy, include subscription, implementation, integration, add-ons, training, support and the internal time to administer it. For build, include discovery, design, development, testing, deployment, hosting, maintenance, security updates and the internal product owner. A build with no maintenance budget is not cheaper; it is deferred cost.
3. How fast do you need it?
A configured SaaS product can go live in weeks. A custom build typically takes months, and the first release should be deliberately small. If the business need is urgent and the process is common, buy. If the need is strategic and can be phased, build.
4. How complex are the integrations?
Count the systems that must exchange data and the direction of each flow. Buying a product that already integrates with your payment provider, messaging channel and accounting system can save more than it costs. Building gives you control over the integration contract, but you own the maintenance when an external API changes.
5. What are the security and privacy obligations?
Customer data, payment data and student or patient records carry obligations. A bought product may have better baseline security than a small custom build, but only if it is configured and monitored. A custom build can be designed around your data model, but it needs threat modelling, access control, logging and testing. Neither option is automatically safer.
6. Who owns the data and the roadmap?
Check the contract. Can you export all records in a usable format? Can you take your configuration and integrations elsewhere? Does the vendor control the release schedule, or do you? Custom software gives you more control, but you also become the vendor, whether or not you have the team for it.
7. What is the exit?
Every option needs an exit plan. For SaaS, that means data export, notice periods and transition support. For custom software, that means source code ownership, documentation, infrastructure access and the ability to hand over to another team.
How the options compare
Buy a standard SaaS product. Best for commodity processes, fast rollout and small teams without engineering capacity. The trade-off is limited fit and dependence on the vendor's roadmap.
Configure a flexible platform. Best when the process is close to standard but needs custom fields, approvals and reports. The trade-off is that configuration can become complex and expensive, and upgrades may break customisations.
Use low-code or workflow tools. Best for internal tools, approvals and integrations between existing systems. The trade-off is governance: without ownership, low-code estates become hard to audit and maintain.
Build custom software. Best when the workflow is distinctive, the integration is complex or the current workaround is costing customers. The trade-off is upfront cost, delivery risk and the need for a long-term owner.
Hybrid. Buy the commodity layer, build the differentiating layer and connect them with a clear API. This is often the most durable answer for an SME.
Decision rules
- If a standard product covers at least 80 percent of the process and the remaining gap is not customer-visible, buy and adapt the process.
- If the gap is customer-visible, recurring and costly, consider building that specific step, not the whole system.
- If you cannot name the internal owner of a custom build for the next two years, do not build.
- If the vendor will not provide a data export or a clear exit clause, do not buy.
- If the three-year cost of buying exceeds the build cost by a wide margin and the process is stable, revisit the build case.
- If the process is still changing every quarter, buy or configure; do not freeze it in custom code yet.
Run a pilot before committing
For a buy decision, run a paid pilot with real data and a real team for four to six weeks. For a build decision, deliver a narrow first release that replaces one manual step, then measure adoption before expanding. A pilot is cheaper than a failed rollout and much cheaper than a rebuild.
Questions for either vendor
- Which requirements are included, and which trigger a change order?
- Who owns the source code, configuration and data?
- How are security updates and dependency upgrades handled?
- What is the support model after launch?
- Can we see a reference implementation with a similar process?
- What does exit look like in practice, not only in the contract?
Metrics for the decision
- Hours per week spent on the workaround.
- Error and rework rate in the current process.
- Time to onboard a new employee onto the process.
- Three-year total cost of ownership for each option.
- Integration count and change frequency.
- Adoption rate after the pilot.
Where this fits in our work
If you are weighing a build, start with Custom Software & Integrations. The pricing page explains how we phase discovery, delivery and support. For the buying side of the same decision, read the dashboard buying guide.
Sources
- OWASP Application Security Verification Standard, accessed 17 November 2026.
- ISO/IEC 27001, Information Security Management Systems, accessed 17 November 2026.
- OWASP API Security Top 10, accessed 17 November 2026.
- DataGuidance, Bangladesh jurisdiction overview, accessed 17 November 2026.
- RFC 9457, Problem Details for HTTP APIs, accessed 17 November 2026.
Responses (0)
No comments yet. Add the first one below.
Leave a response
Your email address will not be published. Required fields are marked *.