The advice is to make the right call. Take the time, gather the information, do not rush a decision you cannot reverse. It is repeated in every leadership course and it is defensible on its own terms, which is exactly why it is dangerous: it names one cost, the cost of being wrong, and it is silent about the other one.
The other one is the cost of being late, and in an operating business it is usually larger. Not occasionally larger. Usually. The reason nobody manages it is that the cost of a wrong decision arrives with a name attached and the cost of a slow decision does not. Somebody owns a bad call. Nobody owns a call that took eleven weeks, because there was never a moment where anyone decided to take eleven weeks. It accumulated, one reasonable request for more analysis at a time.
I have spent my career in corporate turnarounds at Berkshire Hathaway, Illinois Tool Works, and Whirlpool, and four years running a small industrial tank liner manufacturer I bought and later sold to its largest vendor. The single most consistent difference between the businesses that recovered and the ones that ground on was not the quality of the decisions. It was the elapsed time between a problem becoming visible and something changing because of it.
The short version: decision latency carries four distinct costs. The option closes, the delay compounds downstream, the waiting itself consumes capacity, and the organization learns that waiting is safe. Only the first is ever calculated. The fourth is the most expensive and the least visible, and it is the one that turns a slow business into a permanently slow one.
Cost One: The Option Closes
Most operating decisions are not permanently available. They have a window, the window is set by something outside your deliberation, and when it shuts the decision is made for you by default.
A supplier holds a price for a quarter. A skilled operator has an offer and will decide in ten days. A customer is running a sourcing review that closes at the end of the month. A piece of equipment is available used at a fraction of new cost until someone else buys it. In every one of those cases the deliberation continues after the window closes, and the eventual decision is announced into a world where it no longer applies.
What makes this invisible is that the default outcome never gets recorded as a decision. Nobody writes down that the business chose not to hire the operator. The requisition simply aged out and the role was quietly re-scoped. Six months later the capacity problem it would have solved is discussed as a separate issue with no memory of the connection.
The discipline that addresses this costs nothing: when a decision enters the queue, write down the date the option closes and treat that date as harder than the date the analysis is due. If the two conflict, the analysis is what gets cut. That inversion is the whole point. In most organizations the analysis date is the fixed one and the option date is a footnote, which is precisely backward.
And be honest about which decisions genuinely have no window. Some do not, and those should be taken slowly and carefully. They are rarer than the queue implies, because “this one is important enough to get right” is the most socially acceptable sentence available for defending a delay nobody wants to own.
Cost Two: The Delay Compounds Downstream
A decision held for six weeks does not cost six weeks. It costs six weeks plus everything downstream that could not start, and in an operating business the downstream chain is long and the multiplication is not obvious from where the decision sits.
Hypothetical: a layout change is proposed and held for six weeks pending analysis. The layout change was a prerequisite for a scheduling change, which was a prerequisite for a lead-time commitment to a customer, which was a prerequisite for a quote that expires. Six weeks at the top becomes a lost quote at the bottom, and the loss is attributed to pricing rather than to the delay that caused it.
This is why decision latency is so badly underestimated by the people holding the decisions. From inside the deliberation, six weeks is six weeks. From the end of the chain it is a missed season. Nobody is doing the arithmetic across the chain because nobody can see the whole chain, and the person who can see the most of it is usually the most junior person on it.
The structural fix is to require that anything entering the decision queue names what is waiting on it. One line. If the answer is nothing, the item can wait cheaply and should be deprioritized. If the answer names three initiatives and a customer commitment, the item is not a decision, it is a blockage, and it should be treated with the urgency a blockage gets rather than the patience a decision gets.
Most queues, examined this way, sort into a small number of genuine blockages and a large number of items nobody is actually waiting on. That sorting alone frequently recovers more speed than any process change.
Cost Three: The Coordination Tax
Open decisions are not free to hold. They consume attention continuously, and the consumption scales with the number of people who have to stay ready for either outcome.
An unresolved decision means every affected team plans twice, or plans for the more likely outcome and carries the risk of rework, or stops planning and loses the lead time. All three are costly and the third is the most common because it is the only one that does not feel like waste while it is happening. Meanwhile the decision reappears on agendas, gets re-explained to people who joined the conversation late, and generates status requests in both directions.
The tax compounds with the size of the organization, which is why large companies are slow in a way that is not fully explained by bureaucracy. It is not just that decisions take longer to make. It is that holding them open costs more per week when more people are affected, so the same six-week delay is genuinely more expensive at a billion dollars of revenue than at fifteen million, even when the decision is identical.
The practical consequence is that the value of deciding faster rises with scale, exactly when the instinct in most organizations is that scale justifies more deliberation. I have run businesses at both ends of that range and the contrast is stark. At small scale an open decision is uncomfortable and visible. At large scale it is comfortable and invisible, and it quietly bills a dozen teams every week it stays open.
Cost Four: The Organization Learns to Wait
This is the expensive one, and it is the reason decision latency is not a problem you fix once.
People are excellent at inferring what an organization rewards. If decisions reliably take two months, the rational response is to stop bringing decisions forward until you are certain, because bringing something forward early means two months of being asked about it. So proposals arrive later, pre-defended, with the uncertainty already removed and the options already narrowed.
What that produces is a business where the decision queue looks healthy and moves reasonably, while the real latency has migrated upstream into the period before anything is proposed. That period is invisible. It does not appear in any process measurement, because nothing is in the queue yet, and a business can be six months slow by this mechanism while reporting a two-week average decision cycle.
A business can be six months slow while reporting a two-week decision cycle, because the waiting has migrated upstream into the period before anyone proposes anything.
The second-order effect is worse. When proposals arrive pre-defended, they arrive as single options rather than as genuine choices, and a leadership team presented with one option is not deciding, it is ratifying. The organization has adapted so successfully to slow decision-making that it has stopped surfacing decisions at all, and everyone involved is behaving reasonably.
You cannot fix this by speeding up the queue. You fix it by changing what happens to someone who brings an uncertain proposal forward early, which is a behavioral question rather than a process one, and it takes a year rather than a quarter.
Why a Wrong Decision Is Usually Cheaper Than a Late One
Because most operating decisions are partially reversible, and reversibility is the variable that should set your speed rather than importance.
A wrong pricing decision is visible within a quarter and correctable. A wrong scheduling change is visible within a week. A wrong hire is expensive and takes months, which is why hiring genuinely deserves more deliberation. A wrong plant closure is close to permanent. Sort your decisions by reversibility rather than by how consequential they feel, and you will find the great majority sit in the correctable band, where the cost of being wrong is bounded and the cost of being late is not.
That asymmetry is the entire argument. A bounded cost against an unbounded one is not a close call, and treating every decision as though it were the plant closure is how businesses end up deliberating carefully about things that would have been cheaper to try. Harvard Business Review’s treatment of making decisions quickly covers the individual-level mechanics well; the organizational version is harder, because the slowness is distributed and nobody experiences themselves as the cause of it.
The honest counterargument deserves stating. Speed genuinely does degrade quality in a specific band: novel decisions with no precedent, where the organization has no pattern to draw on and the information genuinely changes the answer. Those exist. They are a minority of the queue, and identifying them is a much smaller job than deliberating over everything on the chance that any given item might be one.
How to Put a Number on Your Own Latency
You do not need a study. You need six of your own decisions and about two hours, and the number that comes out is usually persuasive enough on its own that no further argument is required.
- Pick the last six significant decisions your leadership team made. Take them as they come rather than selecting, or you will select the fast ones.
- For each, find the date the issue first became visible to someone in the business, not the date it reached the agenda. Ask the people closest to it, because the gap between those two dates is frequently larger than the gap you are about to measure.
- Record the date something actually changed as a result. Not the date it was decided. The date behavior changed.
- Write the elapsed time and, beside it, what was waiting on the decision and for how long.
- For the three with the clearest downstream chain, estimate the cost of the waiting in whatever unit fits: lost capacity hours, deferred margin, a quote that expired. Label these estimates as estimates when you present them.
- Total it, then divide by six to get an average cost per decision-week. That figure is the price of your current deliberation, and it is the number to put beside any future request for another two weeks of analysis.
Two cautions on the arithmetic. Do not include decisions that were slow for a legitimate external reason, such as waiting on a customer or a regulatory response, because including them invites an easy objection that discredits the whole exercise. And present the estimates as estimates, clearly, because the point of the number is to make a cost visible rather than to win an argument with precision it does not have.
Run this once a year. The number will be uncomfortable the first time and the discomfort is the useful part, because the alternative is a business in which the cost of care is never quantified and the cost of error always is, which guarantees that care wins every argument regardless of the arithmetic.
Where these decisions get made, and on what rhythm, is a structural question rather than a cultural one, and it is the job of the operating cadence sitting underneath all of this.
Monday morning: take the oldest open decision on your list, write down the date it first became visible to anyone, and put that date at the top of the next agenda it appears on.
About the Stagnation Assassin
Todd Hagopian is a Fortune 500 transformation executive who has generated $3B in shareholder value across Berkshire Hathaway, Illinois Tool Works, Whirlpool, and JBT Marel, where he serves as Global VP of Product Strategy and Marketing. Known as The Stagnation Assassin, he is the author of two published books: The Unfair Advantage: Weaponizing the Hypomanic Toolbox and Stagnation Assassin: The Anti-Consultant Manifesto. His blog is published in 15+ languages and read by operators worldwide. Bring him to your stage via his speaking page or connect with him on LinkedIn.
Stagnation Genome scoring. Run the six-decision exercise above and send me what you find: the elapsed times, what was waiting, and where the delay actually sat. I will score where your latency is concentrated and whether it has already migrated upstream. Free, and there is no pitch at the end of it. Request your scoring here.