Find out how strong managers make budget decisions when everything sounds important.
Here is a number that should haunt every manager with a budget to protect: only 9% of companies emerge from a downturn stronger than they entered it. That finding comes from a Harvard Business Schoolstudy of 4,700 public companies across three recessions, led by Ranjay Gulati, Nitin Nohria, and Franz Wohlgezogen. The other 91% either cut too deeply, spent too recklessly, or more commonly — spread their resources evenly across everything to avoid making anyone uncomfortable.
That last group is the most instructive. They didn’t make a bad decision. They avoided making one at all. And if you’ve ever sat through a budget round where every department takes a uniform 10% cut, no exceptions, you’ve seen this pattern firsthand.
Informally, the United States Department of War calls this the “peanut butter” approach: spreading cuts uniformly so that nobody feels singled out. It sounds equitable. But McKinsey’s research tells you how it actually fares: of 238 large companies that announced cost-reduction programs, only 26% sustained results beyond four years. Three-quarters of the effort spent on cutting was, by the companies’ own measures, wasted.
The companies in that successful 9% did something different. They paired operational efficiency with selective, strategic investment in the capabilities that mattered most. They didn’t spend more overall. They spent with sharper intent.
Management thinker Peter Drucker captured the underlying principle decades ago with what he called “organized abandonment.” His test was disarmingly simple: if we weren’t already doing this, knowing what we now know, would we start it today? If the answer is no, the next question is not “let’s commission a review.” It is “when do we stop?”
Building on Drucker and drawing from the behavioral research on how budget decisions go wrong, five questions can serve as a practical filter before any budget commitment:
This is Drucker’s abandonment test. It strips away inertia and forces a forward-looking judgement. Many budget lines survive not because they’re valuable but because they’re familiar.
Jeff Bezos’s distinction between irreversible, high-stakes decisions and ones you can walk back. The first category demands rigorous scrutiny. The second can be decided faster and delegated lower.
Yale’s Shane Frederick has shown that asking this question changes decisions dramatically. A budget line considered in isolation almost always looks worthwhile. Measured against its best alternative, the picture sharpens.
Sunk cost thinking is the most common trap in budget season. Wharton’s Marius Guenzel found that even CEOs are 8–9% less likely to divest from failing acquisitions when their personal sunk costs are high.
Both categories are necessary. But when budgets tighten, the instinct is to protect the operational baseline and starve the strategic bets. The 9% did the opposite.
Apply Question 5. If the skill gap is narrow and your team is strong, training builds lasting internal capacity. If the gap is structural — a capability your team cannot develop in time — the hire is the one-way door worth walking through. The distinction is between a skill gap and a capability gap. Training closes the first; only hiring closes the second.
Apply Question 2. Technology licenses reshape workflows and create dependencies; they’re often one-way doors. If the platform fails, you lose the money and the transition time. People development transfers even if the strategy shifts. Then apply Question 1: if you weren’t already using your current tools, would you adopt them today?
Apply Question 4. Is the revenue push driven by genuine opportunity or by sunk cost thinking (“we’ve already invested so much in this pipeline”)? Then Question 5: a contractor pushes maintains targets; a leadership program builds the judgment that compounds. The 9% protected the latter.
When Steve Jobs returned to Apple in 1997, the company was 90 days from insolvency across 350 products. Jobs cut 70% of the product line, killed the Newton, and posted a $309 million profit within a year. That was organized abandonment at speed. During the 2008 recession, he increased R&D spending by 41.8% while competitors froze. The iPad arrived two years later.
Boeing’s 737 MAX tells the inverse story. Under cost pressure, Boeing modified an existing airframe, relied on a single sensor for a critical flight-control system, and structured deals to avoid pilot retraining. Each decision looked reasonable in isolation. Together they produced $19 billion in damage and cost 346 lives. Boeing treated a one-way door as a two-way door, and the door locked behind them.
Loss aversion makes cutting a $500,000 program feel like losing a million. Arizona State’s Paul Décaire found that “use it or lose it” end-of-year spending generates $0.17 in revenue per dollar, compared to $1.08 at other times. And budget season structurally rewards advocates. It rarely rewards the manager who volunteers to kill their own program because the strategy has moved on.
No budget conversation today escapes the AI question, so it’s worth addressing briefly. The five questions apply to AI spending exactly as they apply to any other line. The complication is that AI is simultaneously depreciating some capabilities (routine drafting, basic analysis) and increasing demand for others (judgement, contextual reasoning, the ability to evaluate machine output). Economist David Oks invokes the Jevons paradox (when making something more efficient leads to using more of it overall) — historically, every leap in software productivity has created more demand for skilled work, because cheaper production unlocks uses that weren’t previously viable.
For budget purposes, Question 5 is the test. If your bottleneck is producing first drafts, AI tools help. If the bottleneck is knowing which draft to pursue and which trade-off to make, the investment that matters is your people. The Jevons paradox suggests you may need both, but build the judgement first.
The best budget decision you make this year is probably not an investment. It is the decision to stop funding something that no longer earns its place: clearly, early, and for reasons you can defend in plain language.
AI will not take your job. It will, however, make it painfully clear which parts of your budget were funding activity rather than outcomes. The leaders who thrive will not be the ones who found more money. They will be the ones who found the nerve to say no.
The hardest part is admitting what that something isn’t.
Thought experiment: Jim Collins once asked: If you inherited $20 million but had only ten years to live, what would you stop doing? Apply this question to your business, and see if your organizational mission can give you the answer that your budget projections cannot.
Every organization's needs are different. With our global learning options, we're ready to help you find the right program to achieve your talent development goals.
