The short answer
Organisations often choose backwards: a product demo impresses, and the business adapts to the tool. Instead, map capabilities, take stock, set weighted criteria, score a shortlist (including "use what we have" and "do nothing") and test with real work before committing. The result is a technology stack (the platforms, applications, infrastructure and tools you run on) that fits together and can be supported over time.
Step 1: map business capabilities
List what the business must do, not which software it wants:
- Take and fulfil orders
- Manage stock across locations
- Bill clients by project
- Onboard employees
- Collect payments by bank transfer and card
- Report to funders or regulators
For each, note how it works today, what goes wrong and its importance. Outcomes make very different options comparable: extending your ERP, buying SaaS or building custom (see build, buy or integrate).
Step 2: take stock of what you already have
Many organisations already pay for capabilities they do not use, especially in Microsoft 365, Google Workspace, their accounting system or ERP. Note systems approaching end of support; vendors publish these dates, as in Microsoft's lifecycle policy pages.
Step 3: set evaluation criteria
Agree and weight criteria before looking at products; the worksheet below lists them. Four need the most care:
- Support lifecycle. Unmaintained versions become a risk: WordPress notes older versions are not maintained with security updates. Active open-source projects publish releases openly, as ERPNext does on GitHub.
- Data location and data protection. Where data is stored and processed, who can access it, and whether your contract gives appropriate safeguards when personal data is processed outside Nigeria, as the Nigeria Data Protection Act 2023 (NDPA) requires (NDPC).
- Resilience to local conditions. Does it keep working through power and connectivity interruptions (offline tolerance, sensible sync, light pages on mobile networks)? Cloud tools depend on your links: plan UPS or inverter cover and a second internet connection alongside the choice.
- Total cost of ownership. Licences or hosting, implementation, integrations, training, administration, upgrades and exit, over three to five years, as estimates in naira (or US dollars where a licence or service is priced in dollars). Note dollar-priced items: exchange-rate movements change their cost in naira.
All ten criteria and what to check
| Criterion | What to check |
|---|---|
| Functional fit | How much works out of the box, and how much needs configuration or custom work |
| Integration | Documented APIs, connectors to your systems, clear rules on which system owns which data |
| Security | MFA, single sign-on, role-based permissions, audit logs, encryption, the vendor's record of fixing vulnerabilities; for custom web apps, how the OWASP Top 10 risks are addressed |
| Support lifecycle | How long versions get security updates and how often you must upgrade. WordPress notes older versions are not maintained with security updates; active open-source projects publish releases openly, as ERPNext does on GitHub |
| Data location and data protection | Where data is stored and processed, who can access it, and whether your contract gives appropriate safeguards when personal data is processed outside Nigeria, as the NDPA requires |
| Resilience to local conditions | Keeps working through power and connectivity interruptions; offline tolerance, sensible sync, light pages on mobile networks |
| Skills and support | People in Nigeria who know it; vendor or partner support in your time zone |
| Accessibility and usability | Accessibility standards (we recommend WCAG 2.2 AA as good practice) and good behaviour on phones, where many of your users will work |
| Total cost of ownership | Licences or hosting, implementation, integrations, training, administration, upgrades and exit, over three to five years, as estimates in naira (or US dollars where priced in dollars); dollar-priced items noted |
| Portability | Exporting all your data in a usable format if you leave |
Step 4: shortlist and score
Score two to four options against the weighted criteria, always including improving what you have. Involve the people who will use and support the system, not only budget approvers, and check references from organisations of similar size and sector.
Step 5: test before you commit
Run a proof of concept with your own data and hardest real scenarios: the unusual order, the multi-currency invoice, the approval with three exceptions. Test integrations and permissions, not just screens, and confirm your team or provider can support the result. Record what you chose, why, what you rejected and what would trigger a revisit; it saves time when people change.
Technology scoring worksheet (weight 1 to 3, score 1 to 5)
- Functional fit for our priority capabilities
- Integration with our existing systems
- Security features and vendor security record
- Support lifecycle and upgrade effort
- Data location, data protection terms and access controls
- Behaviour during power and connectivity interruptions
- Availability of skills and support in Nigeria
- Accessibility and mobile experience
- Five-year total cost of ownership, including dollar-priced items
- Data export and exit options
- Results of the proof of concept
Common mistakes
Choosing by feature count
More features mean more to configure, secure and train. Pick what does your job well.
Ignoring integration
A good product that does not connect creates re-keying and errors.
Following trends
Adopt generative AI and other new tools when they solve a defined problem better than simpler options.
Underestimating change
Training, process redesign and data clean-up often take more effort than the software.
Standardising too late
Every extra tool adds licences, logins and security exposure. Retire overlaps.
Limitations
No method removes uncertainty: pricing changes, products are acquired and requirements evolve. Review your stack annually against the same criteria, and plan replacements before end of support.
Next step
Our IT consulting and advisory service runs this process with you, from capability mapping to vendor scoring and proof of concept. When no product fits, custom software development can build what you need. To choose an implementation partner, read how to choose a technology implementation provider.
Sources and further reading
Product capabilities and guidance change. These are the primary sources this article relies on, checked on the review date above.
- Microsoft Lifecycle Policy, Microsoft Learn
- ERPNext releases, Frappe (GitHub)
- Hardening WordPress, WordPress Developer Resources
- OWASP Top 10, OWASP Foundation
- Nigeria Data Protection Act 2023 and GAID 2025, Nigeria Data Protection Commission
This article is general information, not legal, accounting or security advice for your specific situation. Examples are hypothetical unless stated otherwise.