4.4 Portfolio, New Product Introduction, and Life-Cycle Planning in S&OP
Key Takeaways
- The product/service portfolio review is the first step of the S&OP cycle: it fixes launch dates, discontinuations, and change effectivity before the demand review attaches quantities to them.
- New product introduction (NPI) demand has no history, so planners use historical analogy, market size times share, committed channel demand, or a planning-bill attach rate, and plan it as a low/likely/high scenario.
- Double counting is the classic transition error: holding the outgoing item at its trailing statistical forecast while adding the new item on top inflates family volume by the cannibalized amount.
- In a full phase-in/phase-out with no market growth, family volume stays level and only mix shifts, so transition risk lives in routings, supplier lead times, and ramp resource profiles rather than in volume.
- Obsolescence exposure equals on-hand plus on-order minus remaining demand through end-of-life, valued at standard cost less recovery value; run-out plans and last-time buys are S&OP decisions.
Every S&OP cycle opens with a question that has nothing to do with numbers: which products and services are we planning for this month? The product/service portfolio review answers it, and it runs before the demand review covered in 4.1. Portfolio review owns launches, engineering changes, substitutions, and discontinuations; demand review owns quantities; supply review owns resources. Get that order wrong and the last two meetings argue about a product set nobody has agreed on.
Why Portfolio Comes First
A statistical forecast can only extrapolate items that already have history, and sales intelligence is only usable once everyone agrees which items exist in which period. Portfolio review locks the assumptions the demand plan is built on:
- Launch dates and stage-gate status for each development project
- Planned discontinuations, run-out dates, and engineering-change effectivity
- Substitutions, rebrands, packaging changes, and regulatory-driven changes
- Development load that will consume operations resources: prototype builds, pilot runs, qualification lots
That last bullet is the one candidates forget. New products consume capacity before they generate revenue, so the aggregate supply plan must carry new product introduction (NPI) project load alongside production volume.
Getting NPI Demand Into the Aggregate Plan
A new item has no time series, so exponential smoothing and regression have nothing to work with. Aggregate planners build the number from judgment and analogy instead:
- Historical analogy - take the launch curve of the most similar prior product and scale it by market size or distribution breadth.
- Market size x share - total addressable units multiplied by a realistic launch share.
- Committed channel demand - signed distribution slots, design wins, pre-orders.
- Planning bill or attach rate - when the new item is a variant inside an existing family, forecast the family and give the new item a mix percentage.
Plan NPI as a scenario (low / likely / high) with the assumptions written down, not as a point number. The S&OP decision is rarely the exact volume; it is how much capacity, long-lead material, and cash to commit before the launch signal arrives.
Ramp-Up Curves
A ramp-up is the planned period-by-period climb from first shipment to steady-state rate, limited by tooling, qualification, supplier capacity, and yield learning. Two errors show up in exam stems. The first is a step function straight to full rate in month one. The second is planning the ramp in units only: early units consume far more engineering, quality, and setup time each than mature units, so a ramping item needs its own heavier resource profile in rough-cut checks. Remember too that initial pipeline fill (stocking the channel) is one-time demand - read as sell-through, it makes month three look like a collapse.
Cannibalization and the Double-Counting Trap
Cannibalization is the portion of the new item's volume that comes out of an existing item rather than out of the market. The classic transition failure is double counting: the statistical engine keeps forecasting the outgoing item at its trailing rate while sales adds the new item's forecast on top. Family volume inflates by the cannibalized amount, supply commits capacity and materials for demand that never arrives, and the outgoing item runs straight into a write-off.
The control is to plan the transition at family level: forecast total family volume first, then split that total by mix percentage between the outgoing and incoming item. The split moves; the total should not move because of the transition itself.
Worked Phase-In / Phase-Out
Family "Model X" ships a level 10,000 units per month. The new X-1 replaces the old X over five months with full cannibalization and no market growth:
| Month | Old X (units) | New X-1 (units) | Family total |
|---|---|---|---|
| M1 | 8,000 | 2,000 | 10,000 |
| M2 | 6,000 | 4,000 | 10,000 |
| M3 | 3,500 | 6,500 | 10,000 |
| M4 | 1,200 | 8,800 | 10,000 |
| M5 | 0 | 10,000 | 10,000 |
Aggregate volume never moves, so no new capacity is needed for volume. What does move is mix: different routings, new supplier lead times, ramp-weighted resource profiles, and dual tooling during the overlap.
Now the double-count version. Hold old X at its trailing 8,000 in M2 and add the new item's 4,000 on top and the plan reads 12,000 - twenty percent above the true 10,000, in every month of the overlap.
Run-out arithmetic for the old item is the other half of the transition. Remaining planned old-X demand from M2 forward is 6,000 + 3,500 + 1,200 = 10,700 units. If on-hand plus open supply for X is 12,000 units, the expected residual is 12,000 - 10,700 = 1,300 units; at a $30 standard cost that is $39,000 of write-off exposure. The S&OP decision then becomes explicit: cancel the last replenishment, discount the residual into the run-out, or accept the reserve.
Life-Cycle Stage Drives the Planning Response
| Stage | Forecast error | Capacity strategy | Inventory policy | Margin pressure |
|---|---|---|---|---|
| Introduction | Very high; no history, judgment methods only | Flexible, small-lot, subcontract; avoid dedicated tooling | Small, short-coverage buys; expedite rather than stock deep | High unit price but poor absorption and heavy launch cost |
| Growth | High and usually biased low | Lead strategy: add and qualify capacity, second-source key parts | Higher service target; availability protects share | Improving as scale absorbs overhead |
| Maturity | Lowest; statistical models work well | Level loading, cost focus, cushion trimmed | EOQ and reorder-point discipline, ABC control, leaner safety stock | Price competition; cost reduction defends margin |
| Decline | Rising again; lumpy and intermittent | Release or consolidate capacity, move to shared lines | Run-out plan, stop replenishment, last-time buys for service | Thin; obsolescence risk dominates |
Read the table as a decision aid: the same family gets different capacity, inventory, and service answers depending on where it sits in the life cycle. That is exactly why portfolio review has to state the stage before anyone forecasts the quantity.
Obsolescence on the Supply Side
End-of-life planning is an S&OP output, not a warehouse afterthought. Set the end-of-life (EOL) date and the engineering-change effectivity together, then build a run-out plan that consumes on-hand and open supply of the outgoing item, its unique components, and its packaging. Where a unique component will no longer be produced, the answer is a last-time buy (lifetime buy) covering remaining production plus the service-parts obligation - and minimum order quantities often force you to buy more than the calculation needs.
Size the exposure the same way every time:
Exposure = on-hand + on-order - remaining demand through EOL, valued at standard cost less any recovery value.
Excess and obsolete inventory hits the income statement as a write-down, so those numbers belong in the pre-S&OP financial view where a decision can still change them.
Exam Traps
- Portfolio review is not "the marketing meeting" - it is the step that fixes the item set the demand plan will quantify.
- Replacement volume is not incremental family volume; adding both forecasts is double counting.
- A flat family total during a transition does not mean no action, because the mix shift still moves resource loads and materials.
- Item-level effectivity dates and cutover sequencing belong to master scheduling and engineering change control, not to S&OP.
Demand review spends its entire meeting debating the forecast for a product whose launch engineering quietly pushed out by a quarter, and supply review has reserved no capacity for the pilot builds now on the calendar. Which S&OP step broke down?
A family has shipped a level 10,000 units per month. The statistical model forecasts the outgoing item at its trailing 8,000 units for Month 2, and sales forecasts 4,000 units for the replacement item, which is expected to fully cannibalize the outgoing item with no market growth. What should the Month 2 aggregate family plan be?
Remaining planned demand for an item being phased out is 6,000 + 3,500 + 1,200 units over its last three months. On-hand plus open supply is 12,000 units at a $30 standard cost with no recovery value. What write-off exposure should pre-S&OP see?
A family is in the growth stage of its life cycle: volume is climbing fast and the forecast has come in low three periods running. Which planning response fits the stage?