Learning Loop Canvas

Turn any decision into a falsifiable bet.

The Learning Loop is the rigour layer underneath every business-model decision — the difference between a hypothesis and an opinion is a test plan. Five lines on a single page: the bet you're making, the smallest test that would tell you it's right, the leading signal you'd watch this week, the lagging signal that confirms it next month, and the threshold that pre-commits you to pivot, stop, or double down. Print it. Pin it above your desk. Honour the result.

Download / Print PDF Try it in The Studio → Learn more about this tool →

How to run this

  1. Print or fill digitally — one page, one decision.
  2. Start with the bet: what specifically do you believe will happen?
  3. Pick the smallest test that would tell you the bet is right.
  4. Define one signal you'd watch this week, and one for next month.
  5. Pre-commit the threshold — the line that would change your mind.
  6. Lock in an If-Then commitment so you actually act on the result.

Learning Loop

Eric Ries · Build-Measure-Learn · The Lean Startup · 2011
The Studio · Wade Institute of Entrepreneurship
① THE BET
The Hypothesis
What are you about to commit to? What specifically do you believe will happen if you do? Frame it as cause-and-effect: "We believe that [doing X] will cause [outcome Y]."
② THE TEST
Smallest Experiment
What's the cheapest, fastest experiment that would tell you the bet was right? Interview, fake door, concierge, pre-sell, A/B. Test behaviour, not opinions.
③ LEADING SIGNAL · 7 DAYS
This Week
The first behaviour, response, or count within seven days. "X of N people did Y." Specific sample, specific behaviour — no vanity metrics.
④ LAGGING SIGNAL · 30 DAYS
This Month
The downstream consequence within 30 days that confirms the leading signal was real. Revenue, repeat use, retention, referral — harder to fake.
⑤ THRESHOLD
Pivot · Stop · Double Down
The pre-committed number that would change your mind. Write it BEFORE the test runs. "If fewer than X, we pivot. If above Y, we double down." Without this line, you'll rationalize the result.
⑥ IF-THEN COMMITMENT
If is below by , then I will by .

Decision: __________________________________    Date: _____________    Test starts: ____________________

Common mistake

Building before you've decided what you'd need to learn. Design the loop backwards — learn first, measure second, build last. Most founders skip the design step and run an unfalsifiable test.

What good looks like

A threshold you'd be genuinely uncomfortable committing to in writing. If the line feels safe, it's not the line — pick the number that would actually force your hand.

When to move on

When the loop is one page, the test starts this week, and you can name the result that would make you stop — and mean it.

Method & lineage

The Learning Loop draws on Eric Ries' Build-Measure-Learn cycle (The Lean Startup, 2011) and innovation accounting; Strategyzer's Test Card (Osterwalder) for the "we believe / to verify / we're right if" structure; Karl Popper's falsifiability principle — a hypothesis isn't one if no result would change your mind; and pre-committed kill criteria from decision science. Wade teaches it as the rigour layer underneath every business-model decision.

Want WAiDE to walk you through it?

WAiDE will help you frame the bet, pick the smallest test, name the signals, and pre-commit a threshold you'll actually honour. You'll leave with a one-page test plan and a session report.

Start Learning Loop with WAiDE →

More canvases

Lean Canvas → Effectuation → Theory of Change →
Browse all 24 canvases →