Expand the sleep window by fifteen minutes when efficiency improves
Once sleep efficiency exceeds 85–90% for a week, add fifteen minutes to the window — then repeat.
Key takeaways
- What it is: Once sleep efficiency exceeds 85–90% for a week, add fifteen minutes to the window — then repeat.
- Why it works: Sleep restriction is not a permanent state; it is a catalyst that rebuilds sleep quality, then gets titrated upward as the efficiency improves. Expanding the window by fifteen minutes per week when efficiency stays above the threshold gradually returns the person to adequate total sleep while maintaining the efficiency gains. Moving too fast expands the window before the efficiency is stable, reintroducing fragmentation.
- Evidence: Established clinical or therapeutic practice.
- Avoid: Adding more time to the window the first time you have a good night rather than waiting for a full week of consistent data — one good night is noise; a week of 85%+ efficiency is signal.
Why it works
Sleep restriction is not a permanent state; it is a catalyst that rebuilds sleep quality, then gets titrated upward as the efficiency improves. Expanding the window by fifteen minutes per week when efficiency stays above the threshold gradually returns the person to adequate total sleep while maintaining the efficiency gains. Moving too fast expands the window before the efficiency is stable, reintroducing fragmentation.
How to do it
- 1At the end of each week, calculate sleep efficiency from your logs.
- 2If efficiency is 85% or above for five of seven nights, add fifteen minutes to the sleep window (usually by moving bedtime fifteen minutes earlier).
- 3If efficiency is below 85%, hold the current window or reduce by fifteen minutes and wait another week.
What the evidence says
Clinical practiceThe 85–90% efficiency threshold for window expansion is the standard Spielman titration criterion; the gradual titration approach is the established CBT-I clinical protocol.
Honest caveat: The 85% threshold is a clinical heuristic; some protocols use different thresholds. The key principle is that efficiency must be stable before expanding the window.
- — Spielman, Saskin & Thorpy (1987), treatment of chronic insomnia by restriction of time in bed, Sleep
Common mistake
Adding more time to the window the first time you have a good night rather than waiting for a full week of consistent data — one good night is noise; a week of 85%+ efficiency is signal.
IX Coach automatically calculates your weekly efficiency and either advances your sleep window by fifteen minutes or holds it, removing the weekly decision from a person who is sleep-deprived and not thinking optimally.
Practice this with IX Coach →