The Hypothesis-Based Business Model

Limitations of the static approach

The Business Model Canvas is not wrong — but it is structurally blind to what matters most in Uber’s reality.

No concept of decay or erosion

Everything looks equally “valid” on the canvas.

No differentiation by speed of change

The BMC assumes synchronous stability across all elements.

No representation of uncertainty

Key questions are invisible: What assumptions might fail next? Which elements are experimental vs. core? Which bets depend on autonomous driving becoming real? 
The Business Model Canvas documents what is, not what might break.

Strategy and execution are artificially separated

The Business Model Canvas suggests:
Strategy = design
Execution = later problem

No place for decision logic

Critical strategic issues are missing: Who decides pricing rules? Which decisions are automated? Where must humans override algorithms?

In a dynamic market environment, these blind spots are no longer tolerable.

A business model is a system of hypotheses with different lifespans.

Instead of describing components, we manage hypotheses.
The model is organized into three permanent layers:

• Value Hypotheses – why customers participate and pay.
• Execution Hypotheses – how the model operates, scales and remains visible.
• Decision & Control Hypotheses – how decisions are made, automated and governed.

Each hypothesis has:

• exactly one status

  • Stable – valid, monitored

  • Eroding – still valid, weakening

  • Experimental – deliberate uncertainty

  • Failed – no longer viable

  • Sunset – actively phased out

• observable signals
• an explicit review horizon

Strategy is the management of status transitions —
not the defense of structure.

How this replaces the Business Model Canvas