What needs to be finished first
Relevant functions that perceive undesirable effects (UDEs) are identified and represented in the workshop.
At least one person understands ToC logic and can distinguish between sufficient cause and necessary condition.
What needs to be ready before start
Large whiteboard or Miro board (at least 6 m horizontal), sticky notes in two colors (UDEs, intermediate causes), arrows/markers for causal links, timer, shared definition of UDE.
A facilitator with ToC experience who runs logic checks; three to eight participants from different functions (Sales, Operations, Engineering, Support); a scribe for assumptions and logic checks.
Scope or system boundary; list of known problems from the last quarters; known improvement initiatives that did not work; definition of "undesirable effect" (UDE).
2-6 h
Mark a UDE area on the right of the board (UDEs at top), causal chain to the left (deeper into system). Post example UDE and example causal arrow as template. Turn off phones and set one uninterrupted time block.
The one question this method answers
Which few core causes lie under the system's undesirable effects, and at which leverage points would an intervention resolve the most UDEs at once?
Marker: Phase
| Step | Duration | Action | Hint |
|---|---|---|---|
1Phase 1: Collect UDEs | 30-45 min | Solo sketching 10 min: each person writes at least five UDEs (negative, measurable, experienced effects). Group clustering, merge duplicates, compress to 10-15 UDEs. Formulate each UDE as complete sentence. | UDEs are not causes but observed effects. "Onboarding is bad" is a diagnosis, "Two of five new customers drop out in week 1" is a UDE. |
2Phase 2: Form causal links | 60-90 min | Ask backward for each UDE: what must hold true for this effect to occur? Post hypotheses as intermediate nodes and connect with arrows. Arrows carry causal statements, for example "if A then B". | Every arrow is checked with CLR (Categories of Legitimate Reservation): clarity, existence, causality, sufficiency, predictability. The person who adds arrows must explain each one. |
3Phase 3: Logic check | 30-45 min | Walk bottom to top together. For each arrow ask: are there counterexamples, is cause sufficient on its own or is a missing condition needed? Mark weak arrows with assumption stickers. | Logic checks are the most expensive phase and the real value of the method. Shortcutting it leaves a mind map instead of a CRT. |
4Phase 4: Identify core causes | 20-30 min | Trace paths and mark common roots. A cause that explains several UDEs is a core-cause candidate. Mark the top 1-3 core causes. | If every UDE has its own root, the tree is incomplete or the UDEs are in multiple systems. In the first case, continue connecting, in the second narrow the scope. |
5Phase 5: Interventions and follow-up | 20-30 min | Collect intervention ideas for each core cause. For each idea note which UDEs it would resolve, owner, next step, and review date. | Interventions are validated later with the Future Reality Tree. CRT diagnoses, FRT creates treatment design; both belong together. |
What comes out at the end
CRT diagram (image or Miro frame) plus structured document with UDE list, full causal graph in text, identified core causes, assumption-sticker list, and intervention ideas with owner.
One diagnostic run per entry with date, area, and participants. Do not overwrite follow-up iterations; record changes in the diagram as a delta list at document end.
Compact working template for Current Reality Tree with context, input, output artifacts, and next step.
What is this method used for?
Which question should be answered at the end?
Which data, observations, or materials are available?
Owner, date, success signal.
Concrete filled scenario, fictional example
Concrete filled scenario, fictional example
Area: Customer Success team, 12 participants, 80 active customer accounts.
UDEs (excerpt, 12 total):
Causal excerpt (arrows from bottom to top):
Interventions:
Assumption stickers: three arrows marked "insufficiently substantiated," spike on CRM data planned.
Recognize symptoms and steer against them
Entries such as "process unclear" or "tool missing" are listed as UDEs with no measurable effects.
In phase 1, enforce: UDE needs actor, frequency, and impact. If a diagnosis is written, reformulate as observed symptom.
Arrows connect nodes but no one can state the "if A, then B" sentence.
Demand a logic statement for each arrow and run CLR checks in phase 3. Remove or sticker assumptions on arrows without a logic statement.
All UDEs sit on one chain, no other roots are opened.
Facilitator requires at least two alternative paths for each UDE. Whoever claims there are no alternative paths must justify it.
Nodes contain names or blame language.
Reformulate each node as system behavior. "X does not communicate" becomes "the communication channel between function A and B is not established."
UDE collection and branching happen, but no core causes are identified.
Set realistic time budget in advance (at least 3 h). If not possible, split into two sessions, but complete phase 4 in one of them.
Done signals checkable in under a minute
Go to the profile for purpose, similar methods, and sources or continue to the next method in the catalog.