Colabmo development course A-P04

Learning objective: Separate necessary functions from the roles, tools and sequence used today.

A current-state map helps locate evidence and pain points. It is not the foundation of the future design. If the team starts by redrawing its swimlanes, it can preserve the same handoffs with new names. First list what must be accomplished, without assigning those functions to departments.

For a request-to-decision process, functions might include identifying the requester, understanding the need, determining eligibility, establishing authority, deciding, communicating the outcome and retaining the required evidence. “Finance signs the spreadsheet” combines a function, an actor and a tool. The underlying function may be establishing spending authority, confirming a budget, or something else. Ask the owner which risk the signature controls.

Decompose until each function has an identifiable input, transformation or decision, and output. A function is too broad if no one can explain what evidence would show it happened. It is too detailed for the current exercise if further decomposition cannot change the design choice or the test.

Distinguish logical dependency from historical sequence. A final eligibility decision depends on sufficient eligibility evidence. Identity and service-capacity checks might run independently. That possibility remains a hypothesis until information dependencies and exception behavior are understood. Parallel work is not automatically better if it creates excessive rework.

Use three questions at each step: What outcome becomes impossible without this function? Which other function needs its output? Could a different actor, time, channel or mechanism deliver that output? Retain a function whose necessity is supported. Put an uncertain function in the evidence ledger instead of deleting it by enthusiasm.

Include the failure path. A sound design knows what happens when information is missing, authority is disputed or a request is rejected. Exception handling is part of the service, not a footnote added after the attractive flow is drawn.

Duncker’s candle problem: seeing a new function

Try before revealing the solution. This is Colabmo’s classroom adaptation of the candle problem associated with Karl Duncker’s research on functional fixedness. Imagine a candle, matches, a box of thumbtacks and a vertical cork board above a table. Using those supplies, propose a way to support the candle on the board so that wax would not drip onto the table. Sketch your idea; do not light a candle or build a live-fire demonstration.

Spend two minutes on your own. Record your first interpretation of each object, then list the functions needed: support the candle, attach that support to the board and catch drips. If you already know the answer, explain the shift in reasoning without revealing it to others.

Reveal the solution and reasoning

Empty the thumbtack box, fasten its rear wall to the board and use its upward-facing tray as a support for the candle. The box changes from a container for tacks into a platform and drip tray. The drawing illustrates the conceptual solution; it is not a tested candle holder.

Functional fixedness is the difficulty of seeing a use beyond an object’s familiar role. Describing the box only as “where the tacks belong” can hide properties that enable another function: a flat base, sides and a surface that can be attached.

Transfer the insight to a business process

“The spreadsheet is our approval process” can conceal several functions: collecting evidence, identifying an authorized decision-maker, recording a decision and communicating it. A spreadsheet is one possible mechanism. Each function still needs a credible owner, evidence and an exception path if the mechanism changes.

In W3, record one familiar label you challenged, the function underneath it, an alternative mechanism and the evidence needed to judge that alternative. Distinguish a constraint in the task from a constraint you supplied yourself.

Evidence boundary: this activity illustrates a reasoning obstacle. It does not diagnose intelligence, establish a participant’s competence, or demonstrate that an FPT workshop improves business results. Familiarity with the puzzle affects the activity. Duncker’s work and Adamson’s later repetition are research sources; our simplified prompt is not an exact replication.

Sources: Karl Duncker, On Problem-Solving (1945 English publication); R. E. Adamson, Functional fixedness as related to problem solving: a repetition of three experiments (1952).

Practice

Use W3 to decompose customer onboarding into six to nine functions. Ban department and software names for the first ten minutes. Draw only dependencies you can explain. Mark one dependency as uncertain and name the evidence needed to resolve it.

Check your reasoning

The team says three departments must approve in order because the current form has three boxes. What should you ask next?