- A P&L that looks fine
- What does gross margin compression actually tell you?
- What does SG&A creep tell you?
- Mix shift, the reading almost everyone misses
- But the allocation is done correctly, isn’t it?
- Reading the three lines together
- Where the standard reading fails
- What to do with the reading
- Frequently asked questions
Five and a half points. That is the operating margin a business in the worked example below lost over two years, and it is the only number most boards look at. It is also the least informative number on the page, because five and a half points of operating margin can be produced by at least four different underlying conditions, and the four require entirely different responses.
The reading below is line by line. It is built around one example that runs through the whole article, because a P&L is only interpretable in relation to its own history and a list of general principles will not teach you the skill. If you want the operating sequence that follows the reading, it is in how to turn around a manufacturing plant.
A P&L that looks fine
Hypothetical: the figures below are illustrative and are constructed to demonstrate the reading method. They are not drawn from any business.
Take a manufacturer indexed to 100 units of revenue two years ago.
- Revenue: 100 two years ago, 96 today
- Gross margin: 32% two years ago, 27% today
- SG&A: 22% of revenue two years ago, 22.5% today
- Operating margin: 10% two years ago, 4.5% today
What the board sees
Revenue down 4%, operating margin down 5.5 points. The instinct in the room is that costs have gotten out of control, and the SG&A line gets the attention because it is the line a management team can act on directly without touching the factory.
What the P&L is actually saying
SG&A moved half a point. Gross margin moved five. Ninety percent of the deterioration happened above the gross margin line, which means the problem is in what the business makes and what it charges for it, not in what it spends on running itself. Every hour spent on the SG&A line in this business is an hour spent on a tenth of the problem.
That single comparison, the movement of gross margin against the movement of SG&A, is the first reading and it takes about four minutes. Most underperforming businesses I have looked at get this reading wrong in the same direction, and the reason is that SG&A is where the executive team’s own decisions live and therefore where its attention naturally goes.
What does gross margin compression actually tell you?
Gross margin compression tells you that the gap between what you charge and what it costs you to make has narrowed, and it can be produced by four distinct mechanisms: price erosion, input cost inflation you did not pass through, productivity loss, or mix shift. The line itself cannot tell you which, and that is why it gets misread.
Separating the four
Take realized price per unit by product family across the same period. If price is flat and margin fell, you have a cost or productivity problem. If price fell, you have an erosion problem and the next question is whether it was discount drift or a deliberate competitive response.
Then take material cost as a percentage of revenue. Rising material cost with flat price is a pass-through failure, which is a commercial process problem rather than an operations problem, and it is one of the more common causes of the pattern in the example above.
The productivity check
Labor and overhead as a percentage of cost of goods, compared against output per labor hour on the constraint. When output per hour has fallen while headcount held, you have a genuine productivity loss and it will usually trace to downtime, changeover, or scheduling. That check is in the two-week plant diagnostic.
What does SG&A creep tell you?
SG&A creep tells you that fixed cost has grown faster than revenue, and half a point of it over two years is normal organizational drift rather than a finding. The number worth reading is not the percentage but the absolute dollars against the absolute revenue, because a rising SG&A percentage on falling revenue can occur with no spending increase at all.
Percentage versus dollars
In the example, SG&A went from 22 to 22.5 percent of revenue while revenue fell from 100 to 96. Run the arithmetic: SG&A dollars went from 22.0 to 21.6. Spending actually fell. The percentage rose because the denominator shrank, and an executive team looking at the percentage would spend a quarter attacking a cost base that had already been reduced.
When SG&A is the real finding
SG&A is a genuine finding when the dollars have grown while revenue held or fell, and when the growth concentrates in one or two functions rather than spreading evenly. Even growth across every function is usually wage inflation. Concentrated growth is a decision somebody made, and it can be examined.
Mix shift, the reading almost everyone misses
Mix shift is the mechanism where neither price nor cost moved, but the proportion of your revenue coming from lower-margin products rose, and it is the hardest of the four to see because every individual line item looks unchanged. Nobody made a bad decision. The business simply sold a different blend of the same things.
How to test for it
Take revenue by product family for both periods and recompute the prior period’s gross margin using the current period’s mix. If the recomputed number lands close to today’s actual margin, mix explains the compression and no amount of cost work will recover it. If it lands close to the prior period’s actual margin, mix is not your mechanism and you should go back to price or productivity.
That recomputation takes an afternoon and it is the single highest-value piece of analysis available in a P&L review. The method I use to follow it through to a decision is the 80/20 Matrix of Profitability, which arranges customers against products and allocates cost honestly rather than by volume, and it turns a mix finding into a specific list of things to reprice or stop making.
Why mix shifts happen quietly
Mix shifts are almost always the accumulated result of quoting behavior rather than of strategy. Sales quotes what customers ask for, wins what it wins, and the blend drifts toward whatever is easiest to win, which is reliably the lower-margin end. Nobody in the building experiences this as a decision. The Harvard Business Review work on growth outside the core treats the adjacent version of this problem, and the diagnostic instinct transfers: the drift is legible only when you look at the proportions rather than at the totals.
But the allocation is done correctly, isn’t it?
Your allocation is almost certainly done consistently, which is a different thing from correctly, and the distinction is what makes gross margin by product family unreliable in most manufacturing businesses. Standard costing attaches overhead by volume, so a product that consumes four times the changeover time of its neighbor carries the same overhead per unit.
The objection I get at this point is that the total is right, and it is. The total overhead is fully absorbed and the P&L balances. What is wrong is the distribution, and the distribution is the only part you were going to make decisions with. A perfectly accurate total attached to a fictional per-product breakdown gives you a document that is correct in aggregate and misleading in every specific use.
The second objection is that fixing it requires an activity-based costing project. It does not, at this stage. You need a rough attachment of the four costs that vary by product rather than by volume, which are changeover time, tooling upkeep, inventory carried, and the engineering and service attention consumed. Rough and done in three weeks beats precise and delivered after you have left.
Reading the three lines together
The three readings combine into a diagnosis, and the combination matters more than any single line. Gross margin down with price down is erosion. Gross margin down with price flat and material up is a pass-through failure. Gross margin down with price and material flat is either productivity or mix, and the mix recomputation separates those two in an afternoon.
SG&A enters the diagnosis last and usually as a modifier rather than as a cause. The exception is the business where SG&A dollars have grown by more than gross margin dollars have fallen, which does happen and which is a genuinely different situation requiring a genuinely different response.
In the worked example the reading resolves cleanly. SG&A dollars fell, so cost discipline is not the issue. Revenue fell 4% while margin fell 5 points, which is too large a margin move for a volume effect of that size, so absorption is not the whole story either. That leaves price, material, or mix, and the recomputation would settle it in one afternoon.
Where the standard reading fails
The standard reading of an underperforming P&L starts at the bottom and works up, and it fails because the bottom of the statement is where the smallest and most controllable numbers live. Operating margin is down, so look at operating expense. It is a reflex, it is understandable, and in the example above it points at the tenth of the problem that had already improved.
It fails a second way in businesses with falling revenue, where every ratio moves without anything happening. Fixed cost as a percentage of revenue rises, SG&A as a percentage rises, overhead absorption per unit worsens, and a page full of deteriorating ratios describes a single underlying fact, which is that the denominator got smaller. Reading those ratios as five separate findings produces five separate initiatives aimed at one problem.
The third failure is comparing against budget rather than against history. Budget comparison tells you how good your forecasting was. History comparison tells you what changed in the business. Only one of those is a diagnosis, and in a turnaround the budget was usually built by the people whose assumptions you are trying to test.
What to do with the reading
Convert the reading into exactly one sentence naming the mechanism, then attach the first action to it. Gross margin compression driven by mix shift, so recompute the matrix and identify the families to reprice. Gross margin compression driven by pass-through failure, so rebuild the price adjustment process before touching anything in the factory.
Do not commission a full costing project on the strength of the first reading. The reading is a direction, and the rough version is enough to act on in the first quarter. The precise version is a Territory B problem and it can wait until you are no longer under a clock.
And write the reading down with a date on it, before anyone has an interest in the conclusion. Six months from now the mechanism you identified will be disputed by people with a stake in a different explanation, and a dated document written from the numbers themselves is the version that holds.
Frequently asked questions
Should you look at gross margin or operating margin first?
Gross margin, always. Operating margin is a composite that can move for reasons that cancel each other out. Compare the movement of gross margin against the movement of SG&A over the same period and you have located which side of the business the deterioration is on inside five minutes.
How far back should the comparison go?
Eight quarters at minimum, and by quarter rather than by year. A single year hides direction of travel, and direction is most of what you are trying to establish. Two years of quarters also lets you see whether the deterioration is a step change or a drift, and those have different causes.
What if gross margin by product family is not available?
Ask for revenue by product family instead, which every business has, and pair it with a rough per-family cost attachment you build yourself. That is enough to run the mix recomputation, which is the analysis that matters most, and it does not require finance to build anything new.
Is falling SG&A always good news?
No. Falling SG&A dollars in a business with falling revenue can mean discipline or it can mean that the commercial capability that would have arrested the revenue decline has been cut. Check whether the reduction concentrates in the functions that generate demand before you record it as a win.
A limitation worth naming plainly. This reading assumes a business with a coherent product set and a single reporting entity. In a group with several operating units, or one carrying meaningful intercompany transfer pricing, the consolidated P&L will average conditions that have nothing to do with each other and the reading will describe none of them. Run it separately per unit, or it tells you nothing.
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.
Answer this one before you do anything else: over the last eight quarters, did your gross margin dollars or your SG&A dollars move further? If it is the first and your last three initiatives were aimed at the second, that is the whole finding. Send me the two numbers and I will tell you which mechanism you are looking at.