Before correcting in a delayed system, inventory pending…
Before adding corrective action in a delayed-feedback system, inventory pending actions already taken but not yet producing results to avoid pipeline overfilling and subsequent overshoot.
When This Fires
- Before making corrections in any system with significant feedback delays
- When you've already acted but "nothing is changing yet" — before adding more correction
- In organizational contexts: hiring, marketing campaigns, product changes, culture initiatives
- When managing any system where actions taken today produce results weeks or months later
Common Failure Mode
Increasing correction intensity because the problem persists: "We're still short-staffed, let's post more job listings!" But there are already 15 active job postings in the pipeline, 8 candidates in interview stages, and 3 offers pending. When these all materialize, you'll be over-staffed. The current shortage reflects a hiring delay, not a hiring deficit. More postings will worsen the eventual overshoot.
The Protocol
(1) Before adding any corrective action in a delayed system, stop and inventory the pipeline. (2) List all actions already taken but not yet producing results: postings made, campaigns launched, changes implemented, investments made. (3) For each pending action, estimate: when will this produce visible results? How much effect will it have? (4) Sum the expected effects of pending actions. If the sum would fully correct or overcorrect the current discrepancy → wait. Do not add more corrections. (5) Only add new correction if the sum of pending actions is clearly insufficient to address the discrepancy. (6) When in doubt, under-correct. It's easier to add a small additional correction later than to reverse an overshoot caused by pipeline overfill.