Strategy & Advisory
Build, buy or extend: a practical guide for business software
Off-the-shelf, low-code or fully custom? A clear way to weigh cost, fit, risk and long-term ownership before you commit budget to a new system.
When a process outgrows spreadsheets or an ageing system needs replacing, the first big decision is how to replace it. There are broadly three options: buy a product, extend a platform you already own, or build something custom. Each can be the right answer. The wrong one is expensive to undo.
Option 1: Buy an off-the-shelf product
Buying makes sense when your process is common and not a source of competitive advantage: payroll, accounting, standard ticketing and so on.
- Strengths: fast to start, proven, supported by the vendor
- Watch out for: licence costs that grow with users, limited flexibility, and pressure to change your process to fit the product
Option 2: Extend a platform you already own
Many organisations already pay for Microsoft 365, Dynamics 365 or Power Platform. Extending those can deliver a tailored solution quickly, on infrastructure, security and identity you already manage.
- Strengths: lower cost, familiar tools, integrated with existing data
- Watch out for: licensing implications, and the need for proper governance so the solution stays supportable
Option 3: Build a custom application
Building is right when the process is genuinely distinctive, when nothing on the market fits, or when you need full control over the experience and the data.
- Strengths: exact fit, no per-user licences, you own the code
- Watch out for: higher upfront cost, and the need for a long-term plan for support and change
Five questions to decide
- How distinctive is the process? Common processes favour buying; distinctive ones favour building.
- What do you already own? Existing platforms can make extending the cheapest option.
- What is the total cost over five years? Include licences, support, change and exit costs, not just the build.
- Who will own it afterwards? A solution nobody can support is a liability, however well it is built.
- How quickly do you need value? A phased approach can start with a product or low-code and evolve later.
Make the decision with evidence
Write down the requirements, weight them by importance, and score each option against them, including cost and risk. Run scripted demos so vendors show the same scenarios. The result is a decision you can explain to your board, and a much lower chance of an expensive change of direction later.
