Go-live is an output. A benefit is a change in the business that someone can measure after the balloons are in the bin. If you close the project and walk away, you have delivered hope.
Output is not outcome
A new platform is an output. Fewer hours spent on rework, a shorter cycle time, a revenue change, a risk reduced — those are outcomes. Benefits realisation is the discipline of defining those outcomes early, assigning an owner who outlives the project, and tracking them until they appear or until you admit they will not.
Write benefits as cause → change → measure. 'Improve customer experience' is a mood. 'Reduce average handle time from 9 minutes to 6 within two quarters of go-live, owned by the contact-centre director' is a benefit you can manage. The second version tells you what to instrument and who to bother when the number does not move.
The PM often facilitates the definition. The owner is almost never the PM in year two. If you cannot name the operational leader, you do not have a benefit. You have a slide. A slide cannot receive a benefit. A person can.
Design the handover or design a cliff
Benefits die at handover when operations inherit a system with no process change, no incentive change, and no time to learn. Build the realisation plan into the program or project from the start: training, role design, data for the measure, and a hypercare window with a named owner on both sides.
Baseline the measure before you change anything. 'We think it was bad' is not a baseline. If you cannot measure today, part of the work is to install the measurement. Shipping without a baseline is how everyone argues about success using anecdotes and their favourite customer.
Some benefits are delayed by design — a regulatory benefit that appears at the next audit, a cost-out that appears when a contract ends. Put those on a timeline with review dates. A benefits register that only lists go-live week is a launch checklist wearing a serious name.
When the number does not move
If the output landed and the benefit did not, you have a diagnosis problem, not a cheerleading problem. Was the benefit assumption wrong? Did users not adopt? Did a dependency in another program fail? Is the measure dirty? Each answer implies a different next action: change the process, fund adoption, re-sequence, or stop claiming the benefit.
Do not keep reporting 'on track for benefits' because the project RAG was green. That is how portfolios fund the next cousin of a failed idea. Honesty here is a Business Environment skill and a portfolio skill. It is also basic professionalism.
Disbenefits exist. A faster process that burns out staff, a cost-out that destroys service, a feature that creates a new risk. Log them. A realisation plan that only admits upside is a brochure. Brochures do not help a steerco decide.
Summary & next steps
Define benefits as measurable changes with owners who outlive the project. Baseline, hand over, and keep reviewing after go-live. A green delivery RAG is not a realised benefit.
Next: strategic alignment — how portfolio, program, and project stay connected to what the organisation said it was trying to do.