We looked for the cost of open management loops. Here's what the evidence actually supports.

SaaS marketing loves a convenient statistic. We wanted one too: a rigorous study showing that companies lose some universal X% of revenue because management work stays open — decisions unexecuted, follow-ups stalled, problems that resurface 60 days after being "solved". So we went looking for it. It doesn't exist. This post is about what does exist, where the studies overlap, and the model we think is defensible instead.

1. We searched for X

The claim we wanted to make was simple: "open management loops cost companies X% of revenue." A management loop, in Clio's canonical definition, is a work situation that requires management action, an owner, a due date, a closure criterion, and later verification. When loops stay open, something should leak — time, margin, cash. Surely someone had measured it.

2. X isn't supported

Nobody has. The closest candidates all measure something else:

• Bain's 21% is a productive-power index built partly on self-reported estimates — not revenue, and not open-loop cost.
• McKinsey's $250M per Fortune 500 is presented, by McKinsey itself, as a thought experiment built on headcount and salary assumptions — not an observed loss.
• PwC's "almost 30% higher margins" for top-quintile decision processes is a controlled correlation across 4,701 CEOs — an association, not a causal effect you can promise.

If a vendor tells you a universal percentage, ask where it came from. A competent CFO will.

3. Here's what is supported

The absence of a universal percentage does not mean the problem is small. Five findings hold up well:

• Decision time is large and partly wasted. McKinsey survey respondents (1,259 participants, 91 countries) reported spending 37% of their time on decisions and rated 58% of that time as used ineffectively — arithmetically, roughly 21% of total management time exposed to decision inefficiency. A derived benchmark, not a measured company loss.
• Even successful change captures only part of its value. In a McKinsey survey of 1,034 transformation participants, transformations rated successful captured on average only 67% of their maximum financial benefit, with 55% of the value loss occurring during or after implementation. That is the closest large-sample evidence for "we decided, but never fully converted it into a verified result".
• Organizational drag is material. Bain/EIU estimate ~21% of productive power lost to drag, from audits at 25 multinationals plus a 300+ executive survey — a good order-of-magnitude check, with self-report caveats the authors themselves flag.
• Decision process quality is associated with better financials. PwC's controlled analysis found top-20% decision processes associated with industry-adjusted margins almost 30% higher; Bain separately found ~6 percentage points more total shareholder return in the top quintile of decision effectiveness.
• Closure discipline can causally create value. The strongest causal evidence is Bloom et al.'s randomized controlled trial in 28 Indian textile plants: introducing measurement, defect reviews, follow-up meetings and systematic problem resolution raised productivity 17% in the first year (~$325k/year of estimated profit per plant). Causal there — not a universal elasticity.

4. Here's where the studies overlap

The tempting move is to stack the numbers: Bain's 21% of productive power, Deloitte's 41% of the day spent on non-value work, McKinsey's ~21% of management time. Add them up and you get a monster figure — and a methodological embarrassment. These studies measure partially overlapping phenomena: the wasted meeting is inside the drag index, which is inside the ineffective decision time, which touches the non-value work. Summing them counts the same hour three times.

Benchmarks inform assumptions. Company data determines the estimate.

5. The model we think is defensible

Instead of a top-down percentage, we model Loop Leakage bottom-up, in five ledgers kept deliberately separate:

1. Management capacity — avoidable manager hours × fully loaded cost. Repeated meetings on the same loop, reopened decisions, time hunting for evidence. This is capacity value, not EBITDA, until it demonstrably avoids headcount, overtime or contractors.
2. Delayed contribution margin — revenue at risk × contribution margin × the incremental loss probability caused by the delay. A $1M opportunity delayed 30 days is not "$1M lost"; at 40% margin and 12pp of added loss probability it is ~$48k of expected loss. That version survives a CFO.
3. Rework — rework hours × cost per hour, plus direct spend. Any hour counted here leaves the management ledger.
4. Operational leakage — overtime, expedites, premium freight, penalties, credits, defects, avoidable churn margin. Finance usually already has these numbers.
5. Working capital — the tied-up principal reported separately (cash release, balance-sheet exposure), and its annual carrying cost at the company's rate. "$5M tied up" and "$500k/year of carrying cost at 10%" are different statements; neither is "$5M lost per year".

One rule makes the whole thing credible: every economic event belongs to exactly one primary ledger. A delayed contract can produce lost margin (delay), 30 legal hours redone (rework), 10 committee hours spent discussing it a third time (management), a $20k penalty already inside the margin loss (counted nowhere else), and a $500k receivable pushed out (balance sheet, not revenue loss). No double counting — that discipline is the difference between a business case and a stats wall.

And one separation makes it honest: capacity, P&L and cash are reported apart. We don't convert every recovered hour into EBITDA, or every delayed dollar into lost revenue.

6. Here's what we still don't know

The counterfactual — what would have happened if the loop had closed earlier — is the hardest part, and no external study can supply it for your company. It takes cohorts, historical baselines and matched comparisons built on your own data. That is also the opportunity: instrumented loops generate the explanatory variables almost no company has today (time open, reopen count, decision latency, action completion, closure evidence), and finance already has the outcomes. With enough observed loops, the benchmark the literature lacks can eventually be built empirically — per industry, size, loop type and function. Until then, we treat every scenario as illustrative and label it that way.

In short

The literature strongly supports the claim that open management loops represent material economic leakage. It does not support a universal percentage of revenue — and we won't invent one. A Management Loop tells you what work is still open. Loop Leakage tells you what keeping it open is worth: built from your numbers, validated by the benchmarks, and never the other way around.

Primary sources: McKinsey, Decision making in the age of urgency (2019); McKinsey, Losing from day one (2021) and Maximizing value during implementation (2022); Bain, Time, Talent, Energy (Bain/EIU); Bain Decision Insights; PwC, 28th Annual Global CEO Survey (2025); Bloom, Eifert, Mahajan, McKenzie & Roberts, Does Management Matter? Evidence from India, QJE 2013; Deloitte, 2025 Human Capital Trends; Gartner supply-chain approval rework survey (2026); Microsoft Work Trend Index (2023).