Pattern/Systems Thinking/No. 0378
Feedback Delay
Feedback delay is the time between an action and its observed effects in a feedback loop. In control theory, delayed feedback can cause people or systems to react to outdated conditions. Frequent corrections made before earlier effects appear can produce overshoot or oscillation.
Also called Delayed Feedback
- Evidence
- Well established
- Read
- 6 min
- Links
- 12 connections
- Useful when
- Metrics and incentives · Risk and safety · Running projects · Complex systems and policy
01You've seen this when…
- in life
You raise the thermostat, wait ten minutes, then raise it again. An hour later, you’re opening windows.
- at work
A team changes a signup page on Monday. Conversions take a week to show up, but by Thursday they’ve changed it twice more.
- out in the world
A city approves more buses after complaints about overcrowding. Recruiting drivers takes months, so the council approves another expansion before the first one reaches the timetable.
02The idea
You act, check what happened, and adjust. That simple sequence works well when the consequences arrive quickly. With a delay, the situation you’re reacting to may already be changing because of your previous action.
The delay can sit anywhere along the path. Water takes time to travel through a pipe. A new employee takes time to become productive. A monthly report takes time to reveal a change that has already happened. Several delays can add together.
This matters inside a feedback loop: you compare what you observe with what you want, make a correction, and use the result to choose your next move. A balancing loop aims to close that gap. Delayed consequences can make the correction overshoot it.
You see too little change, add another correction, then discover that both corrections have arrived together. You reverse course. That reversal also takes time. Under the wrong settings, the system swings repeatedly around its target.
The crucial relationship is between how quickly you intervene and how long the system takes to respond. The same delay can be manageable with small, measured adjustments and troublesome with frequent, forceful ones.
03Why it happens
- The signal describes an earlier state. A dashboard may show last week’s conditions. Even a current reading can reflect actions taken long ago, such as a room temperature still responding to an earlier thermostat setting.
- Corrections accumulate out of sight. Orders in transit, trainees awaiting qualification and projects awaiting approval already represent future changes. If each new decision ignores these commitments, several people can independently try to close the same gap.
- Large adjustments magnify the delay. A strong correction may continue working after the target has been reached. The eventual overshoot invites a strong correction in the other direction, creating another delayed response.
- Stocks recover gradually. Changing a flow changes how fast something accumulates or drains. A reservoir, inventory buffer or backlog takes time to reach a desired level. Understanding stocks and flows helps separate an improving rate from a stock that still looks disappointing.
04A worked example
Consider an illustrative shop that wants 20 cases of bottled water on its shelves. Sales rise from 10 to 12 cases a week. Deliveries take two weeks, and two old orders of 10 cases are still due.
In week one, the shop receives 10 cases and sells 12. Stock falls to 18, so the manager orders 12. In week two, stock falls to 16, and the manager orders 18. In week three, the first order of 12 arrives. Stock stays at 16, and the manager orders 24.
What it looks like Increasing orders are failing to restore stock. The manager keeps strengthening the response.
What’s actually going on The first new shipment merely matches weekly sales. The order of 18 is still traveling; when it arrives in week four, stock rises to 22. The order of 24 then pushes stock to 34 in week five. Several attempts to repair the same shortage have stacked up. If the manager now stops ordering abruptly, that decision will reach the shelves two weeks later too.
What would have helped Calculate inventory position: stock on hand plus orders in transit, minus any backorders. Set its target to cover demand during the delivery delay and the desired shelf buffer. This makes incoming corrections visible before placing another order.
John Sterman’s supply-chain experiments documented the broader pattern: participants often under-accounted for outstanding orders, helping produce swings in ordering and inventory. Across a supply chain, those swings can contribute to the bullwhip effect.
05How to spot it
06What to do about it
- Trace the whole response time. Separate the wait before an action begins, the time its effects take to develop, and the lag before you can observe them. Write down a range for each. A quick approval process can still feed a slow physical system.
- Keep a record of pending effects. Put outstanding orders beside current inventory. Put signed offers and trainees beside current staffing. For less predictable changes, record what you expect to happen and when, alongside the uncertainty. Review these commitments before adding another correction.
- Match adjustments to the system’s pace. Use smaller changes when the response is slow or poorly understood. Set a review point long enough to observe a meaningful effect. Continue monitoring in between; urgent safety signals can justify acting sooner. Avoid interpreting every interim reading as a fresh verdict on the intervention.
- Test the timing before increasing the force. A simple spreadsheet or simulation can show what happens if several corrections arrive together, or if the delay doubles. Where possible, add a reliable leading indicator, such as orders dispatched before deliveries arrive. Feedforward control uses incoming information to act ahead of the eventual outcome. Check how accurately that information predicts what follows.
07Where it doesn’t cause swings
A delay alone doesn’t guarantee instability. Long delays can coexist with stable behavior when decision-makers account for pending actions, adjust gently, or use a model of the expected response. Instability depends on the delay, the strength of the correction and the system’s other dynamics.
Waiting also has costs. A fire, a cash shortage or a rapidly growing hazard may demand action before earlier measures can be evaluated. The practical task is to estimate what is already underway while deciding what additional action the risk requires.
An ordinary delivery lead time becomes a feedback problem when its consequences shape later decisions. A parcel taking five days to arrive creates a wait. Reordering it every day because it hasn’t arrived creates repeated intervention.
08Roots
In 1868, James Clerk Maxwell studied the governors used to regulate machinery. A steam engine’s centrifugal governor used spinning weights to help control its speed. Engineers needed to know why some regulating mechanisms settled into steady motion while others produced repeated fluctuations. Maxwell’s mathematical analysis helped establish the study of feedback stability, including the importance of how regulating mechanisms respond over time.
Nearly a century later, Jay Forrester encountered a similar puzzle in business. General Electric’s appliance operations experienced swings in employment that customer demand alone struggled to explain. Forrester, an MIT engineer, traced the internal decisions connecting inventory, production and hiring. Managers could react sensibly to their immediate circumstances while their combined decisions generated instability.
His 1961 book, Industrial Dynamics, brought feedback structures and delays into the study of organizations. Supply-chain simulations later let students experience the problem directly: shelves looked bare while replenishment was already traveling toward them. Through system dynamics and writers such as Donella Meadows, the lesson spread into public policy and environmental management, where delayed responses can contribute to overshoot and collapse.
09How solid is this?
Control theory establishes how delays can destabilize feedback systems under particular response settings. Laboratory supply-chain studies support the behavioral problem of under-accounting for pending effects; a delay alone does not imply overshoot or oscillation.
10Connections
- Countered by Leading vs. Lagging Indicator, Feedforward Control, Safe-to-Fail Experiment
- Can lead toOvershoot and Collapse
- Part ofBullwhip Effect, Balancing Feedback, Feedback Loops, Unintended Consequences
- See alsoStocks and Flows, Second-Order Thinking, Tipping Point, Tragedy of the Commons
+ 2 more in the list
11Origin and sources
An established concept in control theory. Jay W. Forrester made feedback delays central to the study of organizations in Industrial Dynamics (1961).
- [1]Maxwell, J. C. (1868). On governors. Proceedings of the Royal Society of London, 16, 270–283.
- [2]Forrester, J. W. (1961). Industrial Dynamics. MIT Press.
- [3]Sterman, J. D. (1989). Modeling managerial behavior: Misperceptions of feedback in a dynamic decision making experiment. Management Science, 35(3), 321–339.
- [4]Meadows, D. H. (2008). Thinking in Systems: A Primer. Chelsea Green Publishing.
Suggest an edit· Updated 2026-10-02