JH
Jonathan Haber

Shorten your PDSA cycles until feedback is near-daily

More loops beat longer loops: compress the cycle time to accelerate learning.

Key takeaways

  • What it is: More loops beat longer loops: compress the cycle time to accelerate learning.
  • Why it works: The rate of improvement in iterative learning is largely a function of cycle frequency: ten short cycles produce ten data points, while one long cycle produces one. Memory for the specific details of what you tried also degrades with time, making the Study phase of a long cycle rely on reconstruction rather than recall. Shorter cycles keep the data fresh and the lessons current.
  • Evidence: Backed by observational / correlational evidence.
  • Avoid: Setting monthly or quarterly "review cycles" that are really just retrospectives — too slow to act as a learning loop rather than a calendar ritual.

Why it works

The rate of improvement in iterative learning is largely a function of cycle frequency: ten short cycles produce ten data points, while one long cycle produces one. Memory for the specific details of what you tried also degrades with time, making the Study phase of a long cycle rely on reconstruction rather than recall. Shorter cycles keep the data fresh and the lessons current.

How to do it

  1. 1Ask: "What is the shortest window in which I could complete one full loop for this skill?"
  2. 2Design your test so results are visible within that window.
  3. 3Treat each day or week of a new behavior as a complete cycle, not a warm-up for later evaluation.

What the evidence says

Observational

Learning research consistently finds that frequent, spaced feedback accelerates skill acquisition more than equivalent practice without feedback loops. Deming’s own work emphasized fast cycles as the primary mechanism of quality improvement.

Honest caveat: Extremely short cycles may be too brief for some outcomes to materialize; the right cycle length depends on how quickly the effect of an intervention becomes visible.

Common mistake

Setting monthly or quarterly "review cycles" that are really just retrospectives — too slow to act as a learning loop rather than a calendar ritual.

IX Coach tracks cycle-to-cycle changes and alerts you when a pattern is emerging across multiple loops, so the meta-signal of "this keeps not working" is visible before months pass.

Practice this with IX Coach →

Related practices