Skip to content

ZeroBlockers vs Waterfall

Why approve the solution before you have the evidence?

Predictability through phases signed off in order

Waterfall is a phase-gated delivery approach where scope is agreed up front and work moves through sequential approvals. Plan, Analyze, Design, Build, Test, Deploy. Work flows one way through the gates, with a formal change request process for changes.

Waterfall moves work through plan, analyze, design, build, test, and deploy, with a sign-off before each phase starts. Customers first see the product at deploy. Changing it then needs a change request back to an earlier phase.

The solution is fixed before anyone knows it works

For most customer-facing products, the open question is whether a feature will help customers at all. Waterfall agrees the solution first and funds it as a project, so the evidence arrives after most of the money is spent. Replacing the solution then needs a new case and the sponsor’s approval.

It is also the default organizations fall back to when a newer framework leaves funding and approvals open.

Fund the outcome and let the solution change

Who carries the work from idea to satisfied customer

Waterfall

  • Strategy and fundingSponsor and PMO
  • Discover the problem to Test solutionsAnalysts and designers
  • BuildDevelopers
  • ReleaseTest and release
  • Run and measureOperations

ZeroBlockers

“Our projects already handle change.”

Project managers will say that tolerances, change control, and stage boundaries let a project change course. They will add that Finance and regulators need a predictable budget and a clear record of what was approved.

Both points are fair. Change control works well when the change is small or the solution is right. It struggles when the evidence shows the solution itself is wrong, and for customer-facing features that is common. A stage boundary lets leaders stop a project, but swapping to a different solution usually needs a new case. ZeroBlockers keeps a quarterly investment decision and a clear record, and takes the choice of solution out of the approval.

How the approaches differ

The differences explained

Approving a solution does not prove customers want it

The biggest risk with a customer-facing feature is that it does not change what customers do. Most features fail this test. In controlled experiments at Microsoft, Bing, and Google, Ron Kohavi and Stefan Thomke found that most ideas did not improve the metric they targeted, up to 90% at some companies. A detailed, approved specification does nothing about that risk. It fixes the solution before anyone has seen whether customers use it, and the funding is tied to that solution. Upfront design still suits work with known requirements, such as a data migration or a physical installation.

Even the paper usually cited as the origin of Waterfall saw this. In his 1970 paper, Winston Royce called the straight sequential approach risky and recommended iteration, early simulation, and customer involvement. Organizations kept the sequence, dropped that advice, and tied fixed-scope funding to each project.

ZeroBlockers keeps research and delivery together in a Stream Team, a small, persistent team that owns one part of the customer journey through live operation. It tests the riskiest assumption before building out the solution, and often the evidence shows the team can build less.

A better idea should not need a new business case

A feature-based business case ties money to a promised output. When evidence shows the output will not work, the team has to argue for stopping approved work and then win approval for a replacement. Even a well-run change process needs time and the right people in the room.

ZeroBlockers funds the team and the customer scope, and leaves the solution open. The Product Team, the leadership group for the product, funds a scope such as booking and agrees outcome targets. The team can try a different way to raise completed bookings without asking for a new project each time.

Accountability stays. Leaders review the investment every quarter and the outcome evidence every week. Finance approves a budget without a feature list and in return gets an outcome target, a quarterly decision to continue, grow, or stop the funding, and weekly evidence on progress.

Review effort should match the risk of the change

Changing button text and adding a new payment method need different scrutiny. ZeroBlockers release governance sets rules based on the type of change, automated test evidence, the ability to roll back, and the team’s track record. Teams release routine changes within those rules, and risky changes get a closer review.

Keep traceable decisions, due diligence, and legally required approvals. An external certification or a contract milestone still binds the team. The Product Team owns those relationships and plans the work that truly needs a shared schedule.

Teams build the tests, monitoring, and skills to run what they release, and sign-offs come off as those are in place. Start with one part of the product where evidence can change the solution and the release rules can be written down.

Sources and comparison scope

Reviewed September 2026. This page compares ZeroBlockers with Waterfall as documented, with attributed practitioner accounts where available. Descriptions of Waterfall follow its published guidance. Claims about common practice are attributed where they come from a named source. Other statements about what organizations typically do are our analysis. The worked example is our analysis and reports no measured results.