Cascade Pathways

"Every problem starts in one dimension. Smart analysis tracks where it spreads."

Problems don't stay contained. A billing error becomes customer churn. Employee burnout creates quality issues. Understanding cascade pathways lets you see where a problem actually spreads before it's happened everywhere at once.

The Core Principle

Cascade propagation: A problem in one dimension triggers impacts in other dimensions, each with its own severity, evidence, and (often) its own citations.

Why this matters:

• The origin dimension is what's visible. The cascade is what's usually missed.
• Most analysis stops at the origin dimension.
• The full picture lives in how far — and where — the cascade actually reaches.

CASCADE PATHWAY MASTER MAP

                         ┌─────────────────────────────────────────────────┐
                         │              PROBLEM ORIGIN                      │
                         │         (Identify starting dimension)            │
                         └─────────────────────────┬───────────────────────┘
                                                   │
           ┌───────────────────────────────────────┼───────────────────────────────────────┐
           │                                       │                                       │
           ▼                                       ▼                                       ▼
    ┌─────────────┐                        ┌─────────────┐                        ┌─────────────┐
    │  CUSTOMER   │◄──────────────────────►│  EMPLOYEE   │◄──────────────────────►│  REVENUE    │
    │     D1      │                        │     D2      │                        │     D3      │
    └──────┬──────┘                        └──────┬──────┘                        └──────┬──────┘
           │                                      │                                      │
           │         ┌────────────────────────────┼────────────────────────────┐         │
           │         │                            │                            │         │
           │         ▼                            ▼                            ▼         │
           │  ┌─────────────┐              ┌─────────────┐              ┌─────────────┐  │
           └─►│ REGULATORY  │◄────────────►│  QUALITY    │◄────────────►│ OPERATIONAL │◄─┘
              │     D4      │              │     D5      │              │     D6      │
              └─────────────┘              └─────────────┘              └─────────────┘

                         ┌─────────────────────────────────────────────────┐
                         │            CASCADE MULTIPLICATION               │
                         │                                                 │
                         │   Each dimension can trigger any other          │
                         │   Pathway strength varies by problem type       │
                         │   Multiple simultaneous cascades possible       │
                         │                                                 │
                         └─────────────────────────────────────────────────┘

Real Cascade Patterns

Assigning a precise probability to "Customer problems cascade to Revenue 70% of the time" would be a false precision — no case library tracks enough repeated instances of identical problems to support that kind of number honestly. What the real case library can show is the actual, cited cascade notation from real published analyses — read as > for "cascades to" and + for "simultaneously":

Case	Origin	Real Cascade Notation
UC-246 — Starbucks' human-vs-AI bet	Employee (D3)	D3 > D1+D5 > D2 > D6 > D4
UC-302 — Gaming industry AI capstone	Operational (D6)	D6 > D2+D3 > D1+D5 > D4
UC-256 — Private credit blind spot	Revenue (D2)	D2 > D6+D5 > D1+D3 > D4

A pattern worth noticing across all three, and common across the library generally: Regulatory (D4) tends to sit last — it's the dimension that responds to what's already happened elsewhere, rather than driving the cascade itself. Beyond that, the specific path is genuinely case-dependent — origin, sequence, and which dimensions move together all vary with what actually happened. Read cascade notation as a documented, cited trace of one real event, not a statistical prediction rule.

Understanding Cascade Flows: A Real Example

Here's what an actual, cited cascade flow looks like — UC-246, origin D3 (Employee), traced through to D4 (Regulatory):

Notice what's not here: no probability percentage on any arrow. Each step is backed by a citation on the case's own page, not a statistical estimate of how likely it was to happen.

Cascade Depth Analysis

Cascades don't stop at one level. They propagate:

LEVEL 0 (Origin):    Problem occurs in Dimension X
                     │
                     ▼
LEVEL 1 (Primary):   Cascades to Dimensions Y, Z
                     │
                     ▼
LEVEL 2 (Secondary): Y and Z cascade to Dimensions A, B
                     │
                     ▼
LEVEL 3 (Tertiary):  Further propagation...

How Depth Actually Gets Set

There's no formula converting "depth" into a multiplier — depth isn't a dial that inflates a dollar figure. It's a scoping decision, made explicitly before the analysis starts: a CAL script declares DEPTH 3 (or however many levels), which bounds how many cascade hops the analysis traces before stopping. More depth means more dimensions get formally scored and cited — it doesn't mean the case's FETCH score goes up mechanically. A shallow, tightly-scoped analysis and a deep, six-dimension one are scored on the same DRIFT/FETCH terms either way.

Example: Multi-Level Cascade

UC-246 (Starbucks) hits all 6 dimensions across 3 levels — real, cited evidence at every step, no dollar total summed across them (see Scoring Methodology for why summing cascade dollars is a real measurement trap, not just a style choice):

LEVEL 0 (Origin):  D3 Employee — 8,000 "coffeehouse coaches," $500M+ labor investment
                   │
                   ├────────────────────────────┐
                   ▼                            ▼
LEVEL 1:           D1 Customer                  D5 Quality
                   the "third place" experience  "more consistent performance"
                   │                            │
                   └──────────────┬─────────────┘
                                  ▼
LEVEL 2:           D2 Revenue — N. America comps +7.1%
                                  │
                                  ▼
                   D6 Operational — Deep Brew, Green Dot Assist AI
                                  │
                                  ▼
LEVEL 3:           D4 Regulatory — labor relations, ~$1B restructuring

Cascade notation: D3 > D1+D5 > D2 > D6 > D4 · FETCH 2,632 · EXECUTE, High Priority — the real verdict this case actually reached, worked through in full on the Scoring Methodology page.

Cascade Velocity

Some cascades happen immediately. Others take time.

Cascade Type	Velocity	Time to Impact	Example
Immediate	<24 hours	Same day	System outage → Customer impact
Fast	1-7 days	Within week	Quality issue → Customer complaints
Medium	1-4 weeks	Within month	Employee burnout → Quality degradation
Slow	1-3 months	Quarterly	Revenue decline → Hiring freeze
Delayed	3+ months	Long-term	Regulatory → Market reputation

Strategic implication: Fast cascades demand immediate response. Slow cascades allow preventive action.

Mapping Your Cascade

Use this template to trace cascade pathways for any problem — this is what a real cascade actually records, per dimension, evidence-first:

ORIGIN: [Primary Dimension] — score ___/100
   │
   ├── Cascades to: [Dimension] — score ___/100
   │       └── Evidence: _______________________ (cited, not assumed)
   │
   ├── Cascades to: [Dimension] — score ___/100
   │       └── Evidence: _______________________
   │
   └── Cascades to: [Dimension] — score ___/100
           └── Evidence: _______________________

[Repeat for as many levels as the analysis actually traces]

METHODOLOGY: ___/100   (how sound is the reasoning?)
PERFORMANCE: ___/100   (how proven is the outcome so far?)
DRIFT = |METHODOLOGY − PERFORMANCE| = ___

CHIRP: ___   (overall severity, drawing on the dimension scores above)
CONFIDENCE: ___   (0–1, how well-sourced is the whole analysis?)
FETCH = CHIRP × DRIFT × CONFIDENCE = ___

Compare FETCH against THRESHOLD (commonly 1,000) → verdict: ___________

See Scoring Methodology for how each of these terms is actually calculated, worked through step by step.

Preventing Cascade Multiplication

Early Detection Strategies

1. Monitor origin dimension closely — Stop problems before they cascade
2. Watch primary cascade paths — Set alerts for likely targets
3. Track cascade velocity — Fast cascades need immediate action
4. Measure cascade costs — Make unmeasured costs visible

Containment Tactics

Cascade Path	Containment Strategy
Customer → Revenue	Proactive retention outreach, service recovery
Employee → Quality	Temporary quality checks, peer review
Operational → Employee	Resource reallocation, overtime limits
Quality → Customer	Rapid response team, transparency
Regulatory → Revenue	Legal review, compliance audit

Next Steps

📊 Cascade Analysis Guide — Step-by-step cascade mapping process

🎯 Scoring Methodology — Calculate impact including cascades

📖 Case Studies — Real cascade examples with numbers

🔍 Observable Properties — Detect cascade signals early

---

Remember: The problem you can see is just the beginning. Map the cascade to see the full cost. 🪶