TThe Diary of a CEO
← All frameworks
Innovation

Uncharted Innovation Loop

Enter unknown territory, listen hard, fail cheaply, and follow evidence

Difficulty
Advanced
Time to result
~months to results
Steps
6
Confidence
97%

Metcalfe describes innovation as deliberately moving into slightly uncharted territory. A founder begins with incomplete knowledge, then compensates by hiring strong talent, listening deeply, talking to the right people, and learning through repeated attempts. Failure is treated as information within the journey rather than proof that exploration itself was mistaken. Moments of conviction still matter, but they sit inside a loop of experimentation and adjustment: choose a meaningful direction, expose assumptions to capable people, test something concrete, learn from what breaks, and continue when the evidence and purpose justify it. The method is neither blind confidence nor endless research. It uses naivety to challenge convention while relying on listening and feedback to prevent ignorance from becoming stubbornness.

Origin

Metcalfe says Pret was not built from a complete plan. He describes it as hundreds of failures mixed with creative breakthroughs, contrarian choices, talented hires, deep listening, and moments when a risky move felt worth attempting.

Core principles

  • 01New work begins beyond complete certainty
  • 02Naivety can create questions that convention suppresses
  • 03Listening converts uncertainty into learning
  • 04Repeated failure is part of discovery
  • 05Conviction should coexist with adaptation

How to run it

  1. 1

    Choose the frontier

    Select a problem where a genuinely better product, experience, or system could exist but the route is not already known.

    Pro tip State why the problem is worth the uncertainty.

    Watch out Novelty alone is not evidence of customer value.

  2. 2

    Admit incomplete knowledge

    Enter the work without pretending to possess a finished answer. Turn uncertainty into questions and testable assumptions.

    Pro tip Write down what you are least certain about first.

    Watch out Naivety is useful only when paired with learning.

  3. 3

    Recruit and listen

    Bring in people with stronger relevant skills and listen far enough to challenge your preferred answer.

    Pro tip Hire for missing capability rather than similarity to the founder.

    Watch out Do not use expert input merely to validate a decision already made.

  4. 4

    Run a bounded attempt

    Create a concrete experiment that can reveal whether the product, process, or behaviour works better.

    Pro tip Keep the first test small enough that failure remains informative rather than ruinous.

    Watch out Apply appropriate safety, legal, and operational controls.

  5. 5

    Use the failure

    Identify what the attempt disproved or exposed, then change the next version rather than defending the old one.

    Pro tip Record the assumption that changed.

    Watch out Failure without a captured lesson is only repetition.

  6. 6

    Commit or redirect

    Proceed when learning and conviction support the direction; otherwise revise the idea or leave it.

    Pro tip Distinguish courage from attachment by naming the evidence for continuing.

In the wild

Building Pret through repeated discovery

Metcalfe says he had no complete plan and repeatedly moved against prevailing assumptions. He credits talented people, painful listening, failures, and moments of confidence for turning an unclear food concept into a growing business.

A sequence of experiments and learning produced a differentiated company without requiring a fully known route at the start.

Common mistakes

Waiting for complete certainty

Metcalfe argues that creating something new requires entering territory where the answer is not yet fully known.

Confusing ignorance with originality

The loop depends on listening, learning, and capable collaborators rather than celebrating lack of knowledge by itself.

Failing without changing

Repeated attempts become productive only when each one informs the next decision.

Is it for you?

Best for

It is best for product or service innovation where the path is unclear but experiments and expert input are available.

Not ideal for

It is not ideal for safety-critical decisions where unbounded trial and error could cause serious harm or breach regulation.

From the transcript

to create something new you've got to put yourself in slightly uncharted territory

Julian Metcalfe · (33:30)

you've got to be prepared to fail many many times and keep going

Julian Metcalfe · (33:30)

From the episode

Pret & Itsu Founder: How I Built TWO Billion Dollar Brands At The Same Time!: Julian Metcalfe