Project Acceptance Is Not Project Success

Project acceptance is not project success. GPM’s current look back at Zwentendorf shows how far apart those two states can be: Austria’s nuclear power plant was technically complete, but it never entered service after the referendum of 5 November 1978. 50.47 percent voted against commissioning it.

The lesson is not about energy policy. It is about governance: a project can meet scope, schedule and technical acceptance while still missing its intended benefit. At Zwentendorf, the decisive acceptance question was settled only after most capital had already been committed.

Complete does not mean effective

In IT projects, the pattern is less dramatic but common. A migration is technically accepted, yet the business bypasses the new process. A reporting product is live, but nobody uses it as a basis for decisions. A supplier delivers the agreed functionality while the operating organisation has not clarified roles or data quality for running it.

Three layers must therefore be governed separately:

Deliverable: What has demonstrably been built, tested and accepted?
Operational readiness: Who takes over processes, data, support and decisions from day one?
Benefit and acceptance: Which behaviour or metric should change — and who confirms it outside the project team?

In complex software initiatives, I often see the first layer reported precisely while the other two disappear into a closing slide. That merely moves risk beyond go-live.

Test benefits before go-live

The 1978 referendum cannot be equated with an IT steering committee. That is exactly why the case is useful as a mental model: decision-makers must not defer questions of legitimacy, use and long-term viability until the final acceptance.

In practice, each major outcome needs an early benefit owner outside the project team, measurable acceptance criteria and a timetable for non-technical decisions. In a migration reconciliation, the team should not only compare planned against actual data. It should also document which business unit owns the data in the target process and which criteria release its use.

What project leaders should do now

Add three questions to your next steering committee: Who will use the outcome, and from when? How will we measure the expected benefit? Which decision outside the project can still prevent impact? If the answers are only technical, acceptance has progressed further than governance. The relevant metric is not whether the project is complete — but whether its purpose actually starts in operations.