"Should we go agile?" is one of the first questions teams ask when a new project starts. But it's no longer an either-or choice. Most organizations today choose between three broad approaches: predictive (often called waterfall), agile, and hybrid, a combination of the two. Here is what each one means, where the ideas come from, and how to decide.
Predictive ("waterfall"): plan first, then build
A predictive approach plans the full scope, schedule and budget early, then moves through phases in order, such as requirements, design, build, test and release. Change is possible, but it is managed through formal change control.
The approach is often traced back to a 1970 paper by Winston W. Royce, Managing the Development of Large Software Systems. Interestingly, Royce never used the word "waterfall", and he warned that a strictly one-pass, sequential implementation "is risky and invites failure". His paper recommended feedback between phases and doing parts of the work more than once.
Predictive approaches work well when requirements are stable and well understood, when the cost of change late in a project is very high, or when regulation requires formal documentation and sign-off at each stage.
Agile: deliver in small increments and adapt
Agile grew out of the Manifesto for Agile Software Development, written by 17 practitioners in 2001. It sets out four values:
- Individuals and interactions over processes and tools
- Working software over comprehensive documentation
- Customer collaboration over contract negotiation
- Responding to change over following a plan
The manifesto adds: "while there is value in the items on the right, we value the items on the left more." In other words, agile doesn't reject planning or documentation. It prioritizes learning and adapting.
Frameworks such as Scrum put these values into practice through short, fixed-length iterations. In Scrum, each Sprint lasts one month or less and produces a usable increment that can be reviewed with stakeholders.
Agile works well when requirements are uncertain or likely to change, when users can give frequent feedback, and when delivering value early matters more than delivering everything at once.
Hybrid: combine both, deliberately
A hybrid approach uses predictive and agile practices in the same project or program. A common example is a project that plans its budget, regulatory milestones and hardware or vendor deliverables in a predictive way, while the software or customer-facing parts are built in agile iterations.
Hybrid isn't a compromise invented by consultants. PMI's Agile Practice Guide (2017), developed with the Agile Alliance, includes guidance on combining agile and predictive approaches and describes hybrid life cycles as "fit-for-purpose". PMI's current standard, the PMBOK Guide – Eighth Edition, is written for teams that "manage traditional initiatives or lead agile transformations".
What the data says
PMI's Pulse of the Profession 2024 reports that the use of hybrid approaches grew from 20% of projects in 2020 to 31.5% in 2023, an increase PMI describes as 57.5% over three years. The same research found that project teams perform equally well using predictive, hybrid and agile approaches.
The takeaway: no single approach is "the best". What matters is choosing the one that fits the work.
How to choose: a practical starting point
These questions won't make the decision for you, but they point in the right direction:
- How stable are the requirements? Stable and well understood points toward predictive. Unclear or likely to change points toward agile.
- How often can users give feedback? Frequent access to users favors agile iterations.
- What is fixed? Fixed dates, budgets or regulatory milestones usually need predictive planning around them.
- Is the work mixed? If some parts of the project are fixed and others are uncertain, a hybrid approach lets each part use the method that suits it.
- What does your organization need to report? If leadership needs stage gates and forecasts while teams want to iterate, hybrid can give both sides what they need.
If you answered "it depends on the part of the project" more than once, a hybrid approach is worth a serious look.
Sources
- Winston W. Royce, Managing the Development of Large Software Systems (1970); see Winston W. Royce on Wikipedia
- Manifesto for Agile Software Development (2001)
- Ken Schwaber and Jeff Sutherland, The Scrum Guide (November 2020)
- Agile Alliance, Introducing the Agile Practice Guide
- Project Management Institute, PMBOK Guide
- Project Management Institute, Hybrid Project Management: project management embraces the fit-for-purpose approach
- Project Management Institute, Pulse of the Profession 2024: The Future of Project Work
