The Platforming Decision | MIT Product Family Frameworks

The Platforming Decision

Product platforming is a financial decision at its core. It determines how shared architecture translates into compounding returns across a product family. This tool walks you through the decision and lets you run it on your own products.

A platform is a deliberately architected and governed system for delivering valued variety over time. The decision to platform is not made once. It is made component by component, by answering two questions for each one: how much of it is truly shared today, and how much will it need to change tomorrow. The sequence below condenses the MIT Designing Product Families method into seven moves.

1

Frame it financially

Platforming trades a commonality premium today for compounding savings across later variants. Early variants pay for the platform; later variants harvest it. If the family will not produce enough variants over enough time, do not platform.

Meyer & Lehnerd: platform efficiency and effectiveness
2

Plan the family in three layers

Write a product plan (which variants, for which segments, when), a commonality plan (what is shared), and a differentiation plan (where variety creates value). Variety is not the enemy. Unmanaged variety is.

Robertson & Ulrich: the three-plan method
3

Model the coupling

Map how components interact with a Design Structure Matrix. Highly connected components are naturally integral and risky to change. Decoupled clusters are candidates for stable, reusable modules.

Eppinger & Browning: Design Structure Matrix
4

Measure delivered commonality

The Product line Commonality Index scores how common your components actually are, penalizing parts that look shared but differ in size, material, or assembly. Intentions diverge from reality; the F-35 targeted 80 to 90 percent commonality and achieved 30 to 40.

Kota et al.: PCI with f-factor penalties
5

Forecast generational change

The Generational Variety Index rates how much redesign each component will need in the next generation, driven by shifting customer requirements, technology, and regulation. High GVI components are where your future engineering budget will go.

Martin & Ishii (2002): Design for Variety
6

Plot the quadrants and decide

Plotting GVI against PCI sorts every component into one of four strategic positions: platform it, modularize it, standardize it, or unlock it before it becomes a constraint. This is the chart this tool builds for you.

Integrated redesign quadrant analysis
7

Govern the platform

Appoint a platform manager with scope over both the investment and the savings, at least through the second and third variants where divergence emerges. Distinguish acceptable divergence (market change, learning) from unacceptable divergence (sub-optimization, pursuit of uniqueness).

Platform management and the Variant Impact Matrix

Name the product family

A family is a set of related products that could share architecture. The tool is preloaded with a sample four-product razor family so you can see how it works, then replace it with your own.

List the products (N)

Add every product in the family, including planned variants. PCI needs at least two products to mean anything.

List the components (P)

Choose the components that could potentially be standardized across the family: chassis, handle, housing, motor, software stack. Six to twelve is a good working set.

For each component, record how many of your products use it, then score the three f-factors. Each f-factor is the fraction of those products where the attribute is identical. A part used everywhere but in different sizes and materials is not really common, and the index penalizes it accordingly.

Now rate how much redesign each component will need from this generation to the next, using the standard scale. Think about which customer requirements, technologies, and regulations are moving, and which components absorb that movement.

Scale: 0 no change · 1 minor change, no redesign · 3 some redesign · 6 moderate redesign, under half the component · 9 major redesign, most of the component

–Family PCI
–Avg GVI rating
–Products
–Components
A · Valued VarietyHigh change, low commonality. Variety matters here. Modularize it behind stable interfaces so variants can differ cheaply.
B · Confusing CommonalityCommon today but needs to change. Lock-in risk. Plan the unlock before the platform becomes a constraint.
C · Properly PlatformedStable and shared. This is the platform core. Protect it, amortize it, govern changes through a Variant Impact Matrix.
D · Unvalued UniquenessUnique without a reason. Wasted cost. Standardize it onto the platform.

Component decisions

Each row is a recommendation derived from where the component lands. Treat it as the start of the conversation, not the end.

ComponentCommonalityGVIQuadrantRecommended move
Frameworks: Design Structure Matrix (Eppinger & Browning), Product line Commonality Index (Kota et al.), Generational Variety Index (Martin & Ishii 2002), platform planning (Robertson & Ulrich), platform economics (Meyer & Lehnerd). Built on the MIT Designing Product Families curriculum, MIT System Architecture Group, Dr. Bruce Cameron, Director. Tool by Marlon Arboleda. Scores entered here stay in your browser and are not stored or transmitted.