Four months into the first turnaround I owned outright, the monthly numbers still looked exactly like the ones I had inherited. Nothing on the P&L had moved. Everything I could see with my own eyes had. That gap is normal, it lasts about two quarters, and it is where most turnarounds get abandoned by whoever is funding them.
The five metrics below move before the financials do. They are ordered by how early each one responds, so the first is the one you should expect to see inside a fortnight and the last is the one that confirms the others were real. If you want the sequence they are measuring, it is in how to turn around a manufacturing plant.
Why the P&L is the wrong place to look for early evidence
A P&L is a lagging record built on a monthly cycle, which means the earliest an operating change can appear in it is roughly six weeks after it happened, and the earliest it can appear reliably is a quarter after that. Waiting for it is not patience. It is measuring with the wrong instrument.
There is a second problem. Financial statements aggregate, and aggregation cancels. An improvement in one area and a deterioration in another net to nothing on the page while the business underneath is changing shape. The five metrics below are all narrow on purpose, because a narrow measure cannot cancel itself out.
For a turnaround that fixed the product before the marketing, see my podcast episode Chip Bergh’s Levi’s turnaround sequencing.
Metric 1: output per labor hour on the constraint
This is the earliest responder and it moves within about two weeks of a genuine change, because it measures the one operation that governs everything downstream of it. Track it hourly on the floor and weekly at the review. It is the metric that tells you whether the plant’s capacity is actually changing.
What makes it early is that it is not diluted. Plant-wide productivity averages the constraint together with operations that have slack, and the slack absorbs the improvement. Measuring only the constraint removes the dilution, which is why a real change shows up in it before it shows up anywhere else.
Read the direction rather than the level. A number that has climbed across three consecutive weeks is a signal. A number that jumped once and flattened is usually attention rather than capability, and it will decay when your attention moves.
Metric 2: median time for an escalated decision to return
Almost nobody measures this and it is the best single predictor of whether the rest of the turnaround will hold. Log every decision that leaves the floor, take the median return time weekly, and watch it fall. It responds within two to three weeks of any change to thresholds or escalation rules.
Almost nobody measures how long a decision takes to come back, and it is the best single predictor of whether everything else in a turnaround will hold.
Two readings matter. The median tells you the friction level. The tail tells you whether a specific class of decision is stuck, and a long tail with a good median usually means one category of question has no clear owner.
The failure mode to watch for is a median that falls because supervisors stopped escalating rather than because answers got faster. Guard against it by tracking the volume of escalations alongside the time. Falling time with falling volume is not progress, it is withdrawal.
Metric 3: flow-through
Flow-through is the share of an incremental dollar of revenue that reaches operating profit, and it is the metric that tells you whether the cost structure has actually changed or whether you are simply selling more. Compute it quarterly as the change in operating profit divided by the change in revenue.
It is the middle responder, moving over one to two quarters, and it is diagnostic in both directions. Rising flow-through on flat revenue means the operating model is genuinely more efficient. Flat flow-through on rising revenue means you are growing a structure that has not improved, which is a perfectly pleasant condition until volume turns.
Read it against the prior year rather than against a target. A target is a negotiation. The prior year is a fact, and the comparison answers the question the board is actually asking, which is whether this business converts growth better than it did.
Metric 4: premium freight and expedite spend
This is the cheapest early indicator available and the one most businesses already have sitting in an account nobody reviews. It responds within a month because it is a direct cash consequence of planning failures, so when planning improves the spend falls immediately.
Pull it monthly in absolute dollars, not as a percentage, and split it between freight you paid to get material in and freight you paid to get product out. Inbound expedites point at supplier and planning horizon problems. Outbound expedites point at schedule reliability. They are different diseases with the same symptom.
The reason it works as a leading indicator is that nobody manages it, which means it has not been optimized for appearance. It is one of the few numbers in a struggling business that is telling you the truth by accident.
Metric 5: on-time delivery from the original request date
This is the last of the five to move and the one that confirms the others were real, because it can only improve when capacity, decisions, and planning have all improved together. Expect two to three months of lag and measure it from the date the customer originally asked for, never from a reset promise date.
Measured from a reset promise date, on-time delivery reports how good you are at rescheduling. Measured from the original request, it reports whether you are getting better.
Where it gets interesting is in the divergence. When the first four metrics have moved and this one has not after a full quarter, your improvement is real but it is landing somewhere the customer does not experience, which usually means the gains are being consumed by mix or by a scheduling policy nobody has revisited.
| Metric | Responds in | A stall looks like |
| Constraint output per labor hour | 2 weeks | One jump, then flat |
| Median decision return time | 2 to 3 weeks | Falling time with falling volume |
| Flow-through | 1 to 2 quarters | Flat while revenue rises |
| Premium freight and expedites | 1 month | Outbound falls, inbound does not |
| On-time delivery from request date | 2 to 3 months | Flat after the other four moved |
Report all five on one page, monthly, to whoever is funding the turnaround. The page is not for you. It is the instrument that keeps the work funded through the two quarters before the P&L catches up, and losing that argument is the most common way a working turnaround gets stopped.
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.
If flow-through is flat while revenue climbs, the problem is usually mix. A 15-minute 80/20 Portfolio Performance Audit will show you which products are consuming the improvement you just made.