Rolling Wave Planning: Plan Enough, Not Everything
Somebody asks for the plan and means all of it: eighteen months, activity by activity, with dates. The honest position is often that the first four months can be planned to that level, the next six can be described in blocks with reasonable confidence, and the last eight depend on things that will be known in March. Producing the detailed version anyway is not diligence. It is a document that will be rewritten twice and believed in the meantime.
Rolling wave planning is the technique for that situation, and Section 5 of the PMBOK® Guide Eighth Edition lists it among the methods available for building a plan. The idea is unremarkable: plan the near horizon in detail, the far horizon coarsely, and add detail as information arrives. What separates the disciplined version from an excuse is everything that happens around the deferral.
What is actually deferred
The detail waits. Activity-level breakdown, sequencing, specific durations and named resources for work eleven months out are deferred, because producing them now means inventing them and inviting somebody to rely on them.
The commitment does not wait. The overall scope, the cost envelope, the key dates and the assumptions behind them are stated from the beginning, at whatever level of confidence is honest, with the confidence stated. A plan whose later phases have no numbers at all has not deferred detail; it has deferred the whole question, and a sponsor is entitled to notice the difference.
The conditions that justify it
The information has a known arrival date. This is the condition that separates rolling wave planning from procrastination. The surveys complete in March, the regulatory decision lands in June, the pilot ends in week twelve. Deferring detail until a known event is a plan; deferring it until things become clearer is a hope, and hopes do not have re-planning points.
Planning early would cost more than planning late. Where the work is genuinely dependent on what is found, detailed planning now produces a plan that will be rewritten, and the rewriting is the waste. Where the work is well understood, the argument does not apply, and the honest answer to a request for a full plan is to produce one.
The coarse level still supports a decision. Somebody has to be able to fund and govern the project on the basis of the far-horizon plan, which means it needs enough substance to carry an estimate and a risk position. If the later waves cannot be described well enough to be estimated at all, the project has an uncertainty problem that a planning technique will not solve, and that is worth saying out loud early.
Eleven wards, one survey at a time
A hospital trust was refurbishing eleven inpatient wards across two years, one at a time, while the hospital continued to operate. The wards were nominally identical, built in the same decade to the same drawings, and the temptation to plan all eleven in detail from the first was considerable.
The drawings were not to be trusted. Forty years of alterations meant every ward differed in its services, and intrusive survey could only be done once a ward was decanted, which happened about six weeks before its works started. Detailed planning for ward nine, eighteen months out, would have been a work of fiction based on a drawing set known to be wrong.
What the programme did was plan in three depths. The ward in progress and the next one were planned in full: activities, durations, trades, materials ordered. The two after that were planned to package level with budgets and indicative durations. The remainder carried a duration allowance, a cost envelope per ward, and a list of the assumptions each would be tested against, the largest of which was that asbestos would be found in the floor screed in roughly half of them.
The re-planning was scheduled rather than reactive: two weeks after each decant, when the survey results arrived, the next ward moved into full planning and everything behind it shifted up one depth. That meeting was in the diary for two years in advance and it happened eleven times.
The assumption about screed turned out to be close: six of the eleven. Because the allowance had been stated, the three wards that needed it in the first year did not trigger a funding crisis, and when ward seven came back clear, the programme was able to release the allowance rather than spend it elsewhere. The final ward completed nine days later than the original envelope had suggested, against a plan that had never contained a fictional activity.
What keeps it honest
Name the re-planning points in the plan itself. A date, a trigger and an owner, visible to governance from the start. This converts the deferral from something a project manager is doing into something the project has agreed to do, and it means somebody will notice if it does not happen.
Plan the next wave before the current one ends. Teams that wait until a wave is finished discover the next wave needs three weeks of planning they do not have, and the gap shows up as a stall. Running the planning for the next wave while the current one is delivering is the whole rhythm of the technique.
For a PMP® candidate, the practical reading is that deferring detail is legitimate when the information has a known arrival and the re-planning is scheduled, so a scenario about an incomplete plan is asking whether the deferral was designed. A response that demands full detail for work that cannot yet be described will produce a plan nobody can rely on. A structured PMP exam preparation course spends real attention on situations where a plan is deliberately incomplete and defensible.
On a live project, look at the far end of your own plan and ask two questions of it. What would have to be known before this could be planned properly, and when will that be known? If you can answer both, the plan is deferred by design and you can say so with confidence. If you cannot, the later phases are decoration, and somebody is making commitments against them.
Interested in going further?
Defending an incomplete plan in front of a governance group requires knowing exactly why it is incomplete and when that changes. Omega's PMP® Exam Preparation works through planning under uncertainty where the pressure is for a date.
Rolling wave planning is one of the techniques described in the PMBOK® Guide Eighth Edition.
Ad · Amazon affiliate link.
References
A110: The PMBOK 8 Schedule Performance Domain: What It Really Covers
A111: Dates Are Not a Schedule
A112: Critical Path Explained Without the Mystique
A113: Float Explained: How Much Delay Can You Really Absorb?
A114: Milestones vs Activities: Stop Treating Them as the Same Thing
PMP and PMBOK are registered marks of the Project Management Institute, Inc.
