Morphological Analysis

Thinking Tools4.2 Expansive Idea FormationAlso known as Zwicky Box, General Morphological Analysis

Originated by Fritz Zwicky (Morphological Astronomy, 1948) and developed for decision support by Tom Ritchey (Wicked Problems, Social Messes, 2011).

Breaks a problem into its dimensions, lists the possible values of each, and examines the combinations so that no workable configuration goes unseen.

A five-row box for a grocery delivery service, with the combinations that cannot hold together struck

Each row is one dimension of the service and each option one value; a design is one pick per row, so the box holds 243 designs. The opening design has a pair that cannot work together: change picks until no conflict shows, then look for the two that name a workable type. Illustrative.

Vehicle

Promise

Pricing

Handoff

Who drives

  • No conflict in this combination. It survives the cross-check; compare it with the other survivors.Struck: no carrier in the region flies drones, so a drone service means owning the fleet.Struck: gig couriers bring bikes and cars, not licensed drones.Struck: a drone drops the bag and leaves; it cannot wait for a signature.Struck: a weekly shop weighs more than a drone can lift.Struck: partner carriers run fixed routes once a day and cannot promise the hour.Struck: the pass promises unlimited drops while the carrier charges per parcel, so every heavy user loses money.A workable type, the city hour: bikes beat traffic on short runs, and staff riders can be held to the hour.A workable type, the suburban weekly shop: one van route per street a week, paid for by the pass.
  • No conflict in this combination. It survives the cross-check; compare it with the other survivors.Struck: no carrier in the region flies drones, so a drone service means owning the fleet.Struck: gig couriers bring bikes and cars, not licensed drones.Struck: a drone drops the bag and leaves; it cannot wait for a signature.Struck: a weekly shop weighs more than a drone can lift.Struck: partner carriers run fixed routes once a day and cannot promise the hour.Struck: the pass promises unlimited drops while the carrier charges per parcel, so every heavy user loses money.A workable type, the city hour: bikes beat traffic on short runs, and staff riders can be held to the hour.A workable type, the suburban weekly shop: one van route per street a week, paid for by the pass.
  • No conflict in this combination. It survives the cross-check; compare it with the other survivors.Struck: no carrier in the region flies drones, so a drone service means owning the fleet.Struck: gig couriers bring bikes and cars, not licensed drones.Struck: a drone drops the bag and leaves; it cannot wait for a signature.Struck: a weekly shop weighs more than a drone can lift.Struck: partner carriers run fixed routes once a day and cannot promise the hour.Struck: the pass promises unlimited drops while the carrier charges per parcel, so every heavy user loses money.A workable type, the city hour: bikes beat traffic on short runs, and staff riders can be held to the hour.A workable type, the suburban weekly shop: one van route per street a week, paid for by the pass.
  • No conflict in this combination. It survives the cross-check; compare it with the other survivors.Struck: no carrier in the region flies drones, so a drone service means owning the fleet.Struck: gig couriers bring bikes and cars, not licensed drones.Struck: a drone drops the bag and leaves; it cannot wait for a signature.Struck: a weekly shop weighs more than a drone can lift.Struck: partner carriers run fixed routes once a day and cannot promise the hour.Struck: the pass promises unlimited drops while the carrier charges per parcel, so every heavy user loses money.A workable type, the city hour: bikes beat traffic on short runs, and staff riders can be held to the hour.A workable type, the suburban weekly shop: one van route per street a week, paid for by the pass.
  • Struck: no carrier in the region flies drones, so a drone service means owning the fleet. (vehicle: Drone, fleet: Partner carrier)
  • Struck: gig couriers bring bikes and cars, not licensed drones. (vehicle: Drone, fleet: Gig couriers)
  • Struck: a drone drops the bag and leaves; it cannot wait for a signature. (vehicle: Drone, handoff: Signed for in person)
  • Struck: a weekly shop weighs more than a drone can lift. (vehicle: Drone, speed: Weekly slot)
  • Struck: partner carriers run fixed routes once a day and cannot promise the hour. (speed: Within the hour, fleet: Partner carrier)
  • Struck: the pass promises unlimited drops while the carrier charges per parcel, so every heavy user loses money. (price: Monthly pass, fleet: Partner carrier)
  • A workable type, the city hour: bikes beat traffic on short runs, and staff riders can be held to the hour. (vehicle: Cargo bike, speed: Within the hour, fleet: Own staff)
  • A workable type, the suburban weekly shop: one van route per street a week, paid for by the pass. (vehicle: Van, speed: Weekly slot, price: Monthly pass)

An example

A regional grocery chain wanted a delivery service, and every meeting ended in the same two options: vans like the national chains, or a deal with a gig app. The strategy lead set out a box instead. Five dimensions (vehicle, promise, pricing, handoff, and who drives), three values each, 243 designs in all.

The team did not judge 243 designs. It judged pairs: could a drone work with a partner carrier, could a one-hour promise work with a carrier that runs one route a day? Ninety pair checks took an afternoon and struck every design containing a pair that failed. About forty survived, and they fell into three types. One was the van service everyone already knew. The other two nobody had raised: cargo bikes with staff riders promising the hour in the two dense cities, and a weekly suburban van route paid for by a monthly pass. The chain piloted the bikes in one city that spring.

Why it works

A team asked for options names the ones it has seen, and those are usually two or three cells in a space of hundreds. Laying out every dimension and every value makes the whole space visible at once, so the familiar options show up as a small corner of it.

The cross-check is what makes the space usable. Judging 243 whole designs is impossible in a meeting; judging pairs of values is quick, and one bad pair rules out every design that contains it. Five dimensions of three values make ninety pairs, and a few dozen failures strike most of the box. What survives is small enough to compare, and it includes designs nobody proposed because nobody had looked at those combinations.

The thinking behind it

Fritz Zwicky Fritz Zwicky, a Caltech astronomer known for proposing dark matter and neutron stars, developed general morphological analysis and described it in Morphological Astronomy (The Observatory, 1948). The method breaks a problem into its independent dimensions, lists the options on each, and examines every combination. Its premise is that a problem's solutions are a space to be laid out whole, so that no workable combination is missed because no one happened to think of it.

Tom Ritchey Tom Ritchey extended the box to policy and planning problems with computer support, set out in Wicked Problems, Social Messes: Decision Support Modelling with Morphological Analysis (2011). His cross-consistency assessment judges every pair of values for whether they can hold together, then strikes each whole combination that contains a failed pair, which turns a box of thousands into a short list a group can work with.

How to run it

  1. State the problem as a design to choose. A delivery service, a product line, a policy: something made of parts that can vary.
  2. Name the dimensions. Usually five to twelve; each must be independent, so choosing on one does not fix another.
  3. List the values under each. Two to six per dimension, including the ones that seem unlikely; the unlikely ones are where the new designs come from.
  4. Check every pair of values. Ask whether the two can hold together in the same design, and write a reason for each one struck.
  5. Strike every design with a failed pair. By hand for small boxes; a spreadsheet does it in seconds for large ones.
  6. Group the survivors into types. Look for three to five families that differ in kind, and name each.
  7. Compare the types on the criteria that matter. Cost, risk, and fit with the business, as with any short list.

Where it goes wrong

The dimensions overlap. Vehicle and fleet can be one choice in disguise if every vehicle implies its driver, and the box then holds designs that are the same design twice. Test each pair of dimensions by asking whether any value of one forces a value of the other; if it does, merge them.

The cross-check turns into a vote on taste. Pairs get struck because someone dislikes them, and the box shrinks back to the familiar options. Strike a pair only for a stated reason (physics, law, money, capacity) and keep the reasons with the box.

The box is filled and then ignored. The team admires the survivors and picks the van anyway. End the session by choosing one unfamiliar type to test.

When to use something else

  • SCAMPER. When there is one existing product to vary, and a list of prompts will move faster than a full box.
  • TRIZ. When the design is stuck on one contradiction between two parameters and needs a principle to resolve it.
  • Weighted Decision Matrix. When the short list already exists and the job is ranking it against weighted criteria.

Where it sits in SOLVE eX

Go to the source

  • Morphological Astronomy, The Observatory, 1948
  • Wicked Problems, Social Messes: Decision Support Modelling with Morphological Analysis, Springer, 2011

Back to the catalog