Start with friction, not technology
A software project should begin with a meaningful business constraint. Look for work that consumes experienced people, causes repeated errors, delays revenue, frustrates customers or creates a growing operational risk.
Do not begin with a predetermined answer such as an app, a new platform or AI. First understand the cost and consequence of the current process.
Compare opportunities consistently
A simple scoring exercise can stop the loudest request from automatically becoming the next project. Consider each opportunity against the same factors:
- Value: What cost, revenue, capacity or customer outcome could improve?
- Frequency: How often does the problem occur and how many people does it affect?
- Risk: What happens if nothing changes over the next year?
- Feasibility: Is the required information available and can the process be changed?
- Dependencies: Does another system, supplier or decision need to come first?
Choose a measurable first step
The first release should be large enough to prove value but small enough to change direction. Focus on one user group, one process or one valuable outcome. Agree how success will be measured before delivery begins.
Useful measures include time removed from a process, fewer errors, faster customer response, reduced operational risk or increased conversion. Activity such as screens built or features released is not the commercial result.
Treat the first investment as evidence
A focused first project gives the business more than a piece of software. It creates evidence about adoption, value and the organisation’s ability to change. That evidence should determine the next investment.