Your budget said the gross margin would be 650,000. It came in at 570,500. The board wants to know why, and "volume was down and costs were up" is not an answer. A profit bridge is the answer.
Every finance team has sat through this meeting. The month closes, the variance report goes round, and the gross margin is short by some uncomfortable number. Somebody has prepared a slide with the budget in one column and the actual in the next, and a red figure at the bottom.
Then the question comes: why?
What usually follows is a list of things that happened. A customer pushed an order into the next quarter. Steel went up. A big account negotiated a discount. The euro moved. Every one of those statements is probably true, and none of them is an answer, because nobody in the room can tell you which of them accounts for how much. The list adds up to a mood, not to a number.
A profit bridge fixes exactly this. It is not a report you produce because the auditor asks for it. It is the difference between saying "margin was down because of a few things" and saying "margin was down 79,500, of which 40,000 was input cost, 21,667 was lost volume, 13,333 was a worse product mix and 4,500 was net price."
A profit bridge decomposes the gap between budgeted and actual margin into a small set of named causes that sum exactly to the gap. If the parts do not sum to the whole, it is not a bridge. It is a story with numbers attached.
Start at budgeted margin. Finish at actual margin. Between them, walk a series of steps, where each step is one cause and one amount. Drawn as a waterfall it looks like a staircase from the opening bar to the closing bar, which is where the name comes from.
The discipline is arithmetic, not presentation. There is no residual bucket. There is no line called "other". The moment you allow an unexplained remainder, the bridge stops being evidence and becomes decoration, because the remainder is precisely where the thing you failed to understand is hiding.
Most finance teams already produce three of the pieces without calling it a bridge: a sales variance, a purchase price variance, a usage variance. What they usually lack is the guarantee that the pieces are mutually exclusive and jointly exhaustive. Two separate analyses, each individually defensible, can double count the same effect and neither one will look wrong on its own.
For a business that sells physical product, five effects are enough to explain almost any margin gap. Each one is defined so it cannot overlap with the others.
You sold a different total quantity than planned, with everything else held at budget. Measured as the change in total units multiplied by the average budgeted margin per unit across the whole portfolio.
Within that total quantity, the split between products moved. Selling more of a thin product and less of a fat one costs you margin even if the total unit count is identical. Measured at constant total volume, which is the part almost everyone gets wrong.
You realised a different selling price than budgeted. Measured on actual quantities, because a price change only matters on units you actually shipped.
The unit cost of what you sold differed from standard. This is where purchase price variance, usage variance and overhead absorption land. Also measured on actual quantities.
Anything transacted in a currency other than the reporting currency moved because the rate moved, not because anybody sold or bought differently. This has to come out before price and cost are calculated, or it contaminates both.
A services business substitutes rate and utilisation for price and volume, and a distribution business often splits cost into purchase price and landed cost. The structure does not change. What changes is what you call the units.
Here is a deliberately small case, with two products, so every figure can be checked by hand. The numbers are illustrative and do not describe any real business.
| Budget | Qty | Price | Unit cost | Unit margin | Margin |
|---|---|---|---|---|---|
| Product A | 10,000 | 100 | 60 | 40 | 400,000 |
| Product B | 5,000 | 200 | 150 | 50 | 250,000 |
| Total | 15,000 | 43.33 | 650,000 |
Budget. Revenue 2,000,000, cost of sales 1,350,000, gross margin 32.5 percent.
| Actual | Qty | Price | Unit cost | Unit margin | Margin |
|---|---|---|---|---|---|
| Product A | 11,000 | 98 | 63 | 35 | 385,000 |
| Product B | 3,500 | 205 | 152 | 53 | 185,500 |
| Total | 14,500 | 39.34 | 570,500 |
Actual. Revenue 1,795,500, cost of sales 1,225,000, gross margin 31.77 percent.
The gap is 570,500 minus 650,000, so 79,500 down. Note how little the gross margin percentage moved: 32.5 down to 31.77, three quarters of a point. A percentage is a terrible early warning system, which is the second reason to build a bridge.
Total units fell from 15,000 to 14,500. The average budgeted margin per unit across the portfolio is 650,000 divided by 15,000, so 43.33.
Volume effect = (14,500 minus 15,000) multiplied by 43.33 = 21,667 unfavourable.
Had the budgeted mix held, those 14,500 units would have split two thirds to A and one third to B: 9,667 units of A and 4,833 of B. Actual was 11,000 and 3,500. So A over delivered by 1,333 units and B under delivered by exactly the same 1,333, which is what "at constant total volume" means in practice.
Each of those swapped units is valued at its own budgeted unit margin: A at 40, B at 50.
Mix effect = (1,333 multiplied by 40) plus (minus 1,333 multiplied by 50) = 53,333 minus 66,667 = 13,333 unfavourable.
A sold 2 below budget on 11,000 units. B sold 5 above budget on 3,500 units.
Price effect = (11,000 multiplied by minus 2) plus (3,500 multiplied by 5) = minus 22,000 plus 17,500 = 4,500 unfavourable.
A cost 3 more per unit on 11,000 units. B cost 2 more per unit on 3,500 units.
Cost effect = minus 33,000 minus 7,000 = 40,000 unfavourable.
| The bridge | Amount | Share of gap |
|---|---|---|
| Budgeted margin | 650,000 | |
| Volume | (21,667) | 27% |
| Mix | (13,333) | 17% |
| Price | (4,500) | 6% |
| Cost | (40,000) | 50% |
| Actual margin | 570,500 | 100% |
The four effects sum to 79,500, which is the gap exactly. No residual.
Now read what the bridge says, because it is not what the meeting would have concluded. Half the damage is input cost, and that belongs to procurement and to the standard setting process. A quarter is volume. Only 6 percent is price, even though price is the thing sales will be asked about first, and the price story is not even a discount story: product A was discounted, product B was sold above plan, and the two nearly cancelled.
The mix effect is the one that would never have surfaced on its own. The business sold 1,000 more units of A than planned, which every dashboard would have shown as good news, while quietly losing 1,500 units of its higher margin product. Volume up, margin down.
The most common way to get mix wrong is to compare budgeted units against actual units product by product, and call the difference mix. That is not mix. That is volume and mix added together and then labelled as one of them.
In the example above, A was 1,000 units above budget and B was 1,500 units below. Naively multiplying those by unit margins gives 40,000 minus 75,000 = 35,000 unfavourable, which is neither the volume effect nor the mix effect but the sum of them (21,667 plus 13,333 = 35,000). It is arithmetically right and analytically useless, because it fuses a demand question with a portfolio question, and those two go to different people with different remedies.
Mix is only mix if the quantity differences you are pricing sum to zero. In the example, plus 1,333 and minus 1,333. If your product by product differences do not net to zero, you have volume mixed into your mix.
Price moved and quantity moved in the same period. There is an interaction between them: the extra 1,000 units of A were sold at the lower price, so is that shortfall a price effect on new volume or a volume effect at a new price?
Mathematically there is no single correct answer. What matters is that you make a choice, apply it consistently, and never leave the interaction sitting in an unexplained residual. The convention used above is the standard one in management accounting:
That convention puts the interaction inside the price and cost effects, which is where operating managers expect to find it, because it reflects what actually happened on the units that actually moved. The alternative conventions are defensible too. What is not defensible is changing convention between periods, because then your bridge is no longer comparable to last month's and the trend, which is the most valuable thing a bridge produces, is destroyed.
If any part of the business transacts in another currency, the exchange rate has to come out first, on its own line, before price and cost are computed. Otherwise it lands inside them and misattributes the cause completely.
A short illustration. A subsidiary sells 1,000 units at 200 euro. The price in euro was exactly as budgeted, not a cent of discount. The budget was set at 1.10 dollars to the euro; the actual average rate was 1.05.
The gap is 10,000. The price effect is zero. If the bridge is built on translated figures without an FX line, that 10,000 appears as a 4.5 percent price shortfall, and somebody spends the next month investigating a discount that was never given.
The mechanics are straightforward once the principle is accepted. Restate the actual result at budgeted exchange rates. Everything left between budget and that restated figure is real operating performance. Everything between the restated figure and the reported actual is the exchange rate, and it belongs on its own line where treasury, not sales, has to answer for it.
Translating at the closing rate rather than the average rate for the period. Revenue and cost accrue across the whole month, so they belong at the average. Using the closing rate creates an FX effect that is real in the accounts and meaningless as management information.
Three checks, and all three have to pass.
That second check is the one that fails most often in practice, and it fails for a mundane reason: the bridge is built in a spreadsheet fed by an export, while the profit and loss account is produced by the accounting system. The two drift. A late credit note, a reclassified freight cost, a stock write down posted after the export was taken, and by the time anybody notices, the bridge is describing a month that no longer exists in the books.
A profit bridge for one entity with a handful of products is a perfectly reasonable spreadsheet. It stops being reasonable at a predictable set of thresholds, and most groups cross them without deciding to.
None of this is an argument that spreadsheets are bad. It is an argument that the profit bridge specifically is the wrong thing to keep in one, because its entire value rests on tying exactly to the accounts, and a spreadsheet fed by an export has no mechanism to guarantee that.
Niyantrak Consulting builds costing and performance management systems for manufacturing, trading and services businesses, and the profit bridge is usually the first thing we put in place, because it is the report that tells you whether the rest of the costing framework is sound. A bridge that will not tie is a costing system telling you something is wrong upstream.
Niyantrak Budgeting is the software version of that discipline. It holds the budget for a group of companies, takes actuals in, and produces the decomposition across quantity, mix, price, exchange rate and cost as a first class report rather than a afterthought. Multi currency consolidation isolates the FX effect on its own line. It runs alongside whatever ERP you already have, and it runs entirely on your own computer, so no data leaves the building.
A short conversation is usually enough to tell whether the problem is the analysis or the costing framework underneath it. Niyantrak Consulting works with manufacturing, trading and services businesses on exactly this.