Variant Portfolio Lab
A fictional task chair with 11 attributes and 1,166,400 possible combinations, the same order of complexity as the Herman Miller Aeron, which Infor puts at more than 1.2 million. Decide which options to keep, retire or reprice, and watch what each choice does to volume, cost and the variant space.
Every option, volume, price and cost here is invented. The decisions are the kind I worked on at Herman Miller: which options stay orderable, which retire and how pricing changes, seen from the configuration side, where each option's real running cost shows up. Background: SKU rationalization, zombie variants, combinatorial explosion.
| Option | Share of chairs | Chairs a year | Price adder | Cost to keep a year | Decision | Why |
|---|
Configure
This chair
Configuration code
Price
How common is it?
The chair
Model K has 11 attributes with 45 values between them. Multiplied out, that is 1,166,400 combinations. Five constraints remove the ones that can't be built, leaving 442,260 valid configurations. Nobody builds them all; each chair is assembled to order from shared parts.
What each option is worth, and what it costs
Every chair earns a base contribution. Options with a price adder earn extra contribution on top. Against that, every option offered has a yearly cost to keep: its column in every price book, its configuration rules and test configurations, its swatches and renders in dealer tools, minimum order quantities and slow stock. Low-volume colours and casters carry the highest cost to keep, because those costs don't shrink with volume.
Retire
A retired option's buyers either accept the nearest alternative, the most popular choice in that attribute, or walk away. The substitution slider sets how many accept. Retiring saves the option's cost to keep and shrinks the variant space.
Reprice
Raising an option's adder earns more on every chair that keeps it, and sends some buyers to the next most popular choice. The price sensitivity slider sets how many.
The judgment call
Apply the suggested plan, then drag substitution down to 70%. The same plan that adds money at 90% barely breaks even, because more buyers walk away. Every real rationalization turns on that question: will customers accept the nearest alternative? Order history, dealer conversations and strategic accounts answer it, not the spreadsheet. That's why two low-volume options in the plan are kept on purpose.