Statistics learn from the past. A new product series has no past, a phase out ends one, and a large tender can double a month without warning. These changes need their own step in the cycle: the portfolio review.
Four kinds of change
Product management owns the portfolio review and brings four kinds of change:
- New products: launch date confirmed, ramp up curve agreed, volume per month for the first year.
- Phase outs: last order date, remaining stock and material, plan to use it up.
- Large projects: probability and expected order date, entered as a weighted volume or as a scenario.
- Price and range changes: expected effect on volume and mix, with an owner.
The output is a short list of changes per family that goes into the demand review a week later.
A ramp up and a phase out in one plan
In the pump example, a new compact pump series replaces 40 variants of the old one in September 2027. The family total stays in the plan, around 2,200 to 2,500 units per month. What changes is the split:
- July and August: 100 percent old series. The last order date of 31 August is announced in July.
- September: 60 percent old, 40 percent new.
- October: 20 percent old, from stock only, and 80 percent new.
- November: 100 percent new. Remaining old stock serves service and spare parts.
The portfolio review decides the split. The master schedule turns it into items. Purchasing knows when to stop ordering material for the old series and how much to order for the new one.
Without this step, the statistical forecast keeps predicting the old series in November, and the plant builds pumps nobody will order.
Projects with probabilities
A water utility may order six engineered systems in May, with a probability of 50 percent. Putting three systems into the plan is wrong in both directions: either six come or none.
Two better options:
- Keep the project out of the base plan and prepare a scenario with its effect on engineering and assembly hours.
- Reserve capacity for engineered systems per month and fill the reservation with confirmed projects.
The weighted volume, here three systems, is useful as an indicator for the pipeline, not as a production plan.
Keep it short
A portfolio review takes one to two hours per month. It works with a simple list: type, family, description, start date, volume, probability, owner, status. Changes without an owner are not entered.
The list also protects the demand planner. When sales asks why the forecast ignored the new series, the answer is in the list: either it was entered, or nobody reported it.
The link to strategy
Over time the portfolio review becomes the bridge between product strategy and operations. Launch dates, ramp up curves and phase outs are no longer surprises for the plant. They are part of the plan 18 months ahead, and capacity, materials and inventory follow them.