Alternative to Fixed-Price Contracts in Software Development

Overwhelming evidence shows that fixed-price, fixed-scope contracts fail in software development. I won't rehash the reasons here, as they've been discussed much better elsewhere. Fixed-price contracts favor neither client nor vendor.

Software development is not like a construction project, where there is a fixed start and end. Software development should be seen as product development. Software needs to be constantly modified based on the feedback from stakeholders and the ever-evolving pressures of the industry. It's more like gardening rather than construction.

The contracting model should therefore be like gardening. The most ideal model for software development is a managed service, similar to the contracts you probably already have with law firms or accounting firms. You need to hire skilled and caring gardeners to make sure your software is continuously able to serve its purpose despite changing business and technology needs.

Note that I'm not advocating for staff-augmentation either, which puts the burden of management with the client, and limits the skill sets available to the client.

The advantages of a managed service:

• Aligned to long-term maintainability of the software. Fixed-price encourages short-term delivery of software that meets written specs but will be expensive for the client to adopt to future requirements.
• Allows for quick changes in features, no need for tedious change request process.
• Vendor can assign the people special skills to implement specific tasks at different times, and is overseen by the vendor's seniors (not the case with staff-augmentation).
• Technology is updated appropriately, you're not stuck with the legacy technology that was specified in a fixed-price contract.
• No need to spend excessive time and effort in detailing requirements. Product team can instead iterate requirements with the development team.