Research Note

What decision paralysis actually costs

By James Carter · July 2026

If someone has quoted you a dollar figure for what indecision costs your company, ask where the number came from. We went looking for the primary source behind the figures circulating in this category. There isn’t one. Every version we could trace ran back to a consultancy’s marketing page, a statistics roundup citing another roundup, or nothing at all.

That is worth saying plainly at the top of a note about cost, because the honest version of this subject starts with what the research does not establish.

What independent research does agree on

Four bodies of work, built with different methods in different decades on different populations, converge on two findings.

Baum and Wally (Strategic Management Journal, 2003) collected measures from 318 CEOs between 1996 and 2000 and modeled strategic decision speed against subsequent firm performance. Faster strategic decision-making predicted later growth and profitability.

Eisenhardt (Academy of Management Journal, 1989) studied eight microcomputer firms inductively and found the opposite of the intuitive result: the executive teams that decided fastest drew on more information and generated more alternatives than the slow ones, not fewer. Fast decision-making tracked with stronger performance.

Sull, Homkes and Sull (Harvard Business Review, 2015) surveyed 7,600 managers across 262 companies. They were not asking about decision speed at all, but they documented the machinery around it: when senior executives insist on making the important calls themselves, middle managers learn to escalate conflicts rather than resolve them, and over time lose the capacity to work things out on their own.

PwC (Pulse Survey, May 2025) asked 678 US executives at public and private companies across six sectors, and 57 percent said they were missing opportunities because they could not make decisions quickly enough. Roughly a third of that sample sat in the C-suite proper, with board directors and functional chiefs making up the rest.

The agreement is the argument here, and it holds because the methods do not overlap. One is a four-year longitudinal model of CEO-reported speed. One is inductive case work in a single industry nearly forty years ago. One is a large-sample manager survey. One is a point-in-time survey of American executives. None was designed to confirm the others. Most articles you will read on this topic cite a single study — usually the author’s own — which is why the convergence rarely gets made.

What none of them do is price it. Not one establishes what paralysis costs a specific company.

On segment. Among the companies Sull’s team reached, median annual sales were around $430 million and average headcount around 6,000, concentrated in financial services, information technology, telecommunications and oil and gas, with a third based in emerging markets. The revenue median sits inside the $50M–$1B band. The headcount and the sector mix do not describe most mid-market companies. Baum and Wally’s data ends in 2000. Eisenhardt’s work covers eight firms in one industry and is framed as propositions rather than tested hypotheses. PwC’s sample is the closest to you in nationality and seniority and the furthest from you in rigor, being a single self-report item with no performance outcome attached. Use all of it to aim. None of it to conclude.

A warning about that last number, from the same source. PwC ran the survey again roughly a year later. In the follow-up, 67 percent of executives reported being ahead of competitors on speed of decision-making and execution, and 90 percent said their company was in a stronger position than two years earlier. PwC itself describes this as a sharp shift from the May 2025 result. Executive decision-making did not transform in twelve months — what moved was sentiment. This is the clearest available demonstration of why self-reported decision speed cannot carry weight in an argument about cost, and we are citing the 57 percent while telling you exactly how soft it is.

Where the cost actually lands

The research names the symptom: slow decisions track with worse performance. The instinct that follows is to decide faster. That instinct is where most executive teams lose the next two years, because paralysis is not slowness. Slowness is what you observe. Paralysis is a specific broken discipline, and its cost does not land where you are looking for it.

It lands in three places, none of which appear on a P&L.

Resource lock-up

Undecided means uncommitted, and uncommitted resources stay exactly where they already are. In Sull’s survey, eight in ten managers said their companies fail to exit declining businesses or kill unsuccessful initiatives quickly enough, and only 11 percent believed all of their company’s strategic priorities had the funding and people they needed. Read those two findings together. The cost of the decision you did not make is paid by the initiative that was already right and never got resourced. It is invisible because nothing invoices you for it.

Escalation, and it compounds

Sull’s team documented the pattern directly: top-down intervention teaches middle managers to escalate instead of resolve, and erodes their decision-making capability over time. Our reading of that finding — and we flag it as our reading rather than something the study measured — is that the cost is not linear. Each decision that routes back to you makes the next one more likely to route back. What you are paying for is not a week of delay. It is the slope.

Re-litigation

Escalation of commitment has been studied for more than three decades and holds up under meta-analysis (Sleesman and colleagues, 2012). A decision that was never genuinely made is the one most exposed to being reopened, and reopening is not a cost you pay once.

Why more information will not fix it

Underneath all three sits the thing that makes paralysis feel unfixable from the inside. The team believes it is missing information. Eisenhardt’s finding cuts straight across that belief: the fastest deciders consumed more information, not less. Data was not their constraint.

Here is the claim we make, marked clearly as interpretation rather than established finding. If more information does not produce faster decisions, then “we need more data” is not a diagnosis. It is the most respectable available reason not to commit. The actual variable is whether the team is willing to live with a call it might be wrong about. That is a discipline, not a process.

That claim should be testable, so here is the test. If the problem were process, a better process would fix it — and most executive teams reading this have already installed one. If the problem is the discipline, you would predict something specific instead: the team can name the decision, name the owner, and still watch it come apart, because it gets reopened the following week by someone who sat in the room and did not object. If your team makes a call and it survives seven days without being relitigated, the Decision discipline is intact and your execution problem is somewhere else entirely. That is a falsifiable claim, and it is why treating “execution” as one undifferentiated problem produces two years of expensive nothing.

If you want a number, build your own

Illustrative calculation — not a research finding

Eight people on the leadership team. One significant decision reopened per week. Ninety minutes of the team’s time on each occasion, counting the meeting and the side conversations that precede it. That is 12 person-hours per reopening and roughly 600 hours a year of the most expensive time in the company, spent re-deciding things that were already decided. Multiply by your own loaded rate.

The figure will be uncomfortable. It is also the smallest part of the cost, because the hours are the only part you can count. The initiative that never got funded does not show up in that arithmetic at all.

Which one broke

Which element broke is a diagnostic question rather than a philosophical one. Four disciplines run off a central set of operating priorities, and execution fails when one of them goes.

The Flag Model sets out all five and what rebuilding each one requires. Most CEOs can feel which one is theirs within about ninety seconds of reading the list. Proving it takes longer.

Sources

All primary. Every figure above was verified against the original study or article text.

“James helped us turn seven groups of rivals into one leadership team. By the time US Holdings became Eagle Manufacturing Group and I moved from COO to CEO, we were no longer seven companies protecting our own territory — we were one company working toward the same outcome.”

Ronn Page · former CEO, Eagle Manufacturing Group

As featured in CNN CNN Money Business Insider Client results →

About the author

James Carter

Founder of Be Legendary and creator of the Flag Model™. Twenty-five years inside executive teams; co-author alongside Stephen Covey, Ken Blanchard, Deepak Chopra & Brian Tracy, and featured on CNN and in Business Insider. More about James →

See where your team breaks first.

A Calibration Call is 15 minutes — you leave with a concrete read on your team, whether or not we work together. It's a calibration, not a pitch.

Book a Calibration Call →

Not ready to talk? Start free

Take the free Break-Point Self-Assessment

See which of the five components your team is most likely to lose first — in ~4 minutes. No email required to see your result.

Start the assessment →

Field notes, by email

Straight thinking on executive-team execution.

One short note, roughly monthly, on the disciplines that decide whether a leadership team executes. No fluff, no pitch. Unsubscribe anytime.