Price volume mix analysis

Take the simplest case first, one product, one segment. Call the earlier period (last year, budget, whatever your baseline is) period 0, and the period you are explaining period 1. You have four inputs for each period: volume V, price P, variable cost per unit C, and fixed cost F. Profit in either period is V(P − C) − F. The question a bridge answers is: of the total change ΔProfit = Profit₁ − Profit₀, how much is attributable to volume moving, how much to price moving, how much to unit cost moving, and how much to fixed cost moving, as four numbers that sum to exactly ΔProfit, with nothing left over?

How is the profit bridge calculated?

There's more than one defensible way to slice this. The moment you hold some variables at old values and others at new values, you're making a choice. A different choice gives a slightly different split. So state your convention once and use it consistently, the same habit the Finance track insists on when you construct EBIT. The convention used in every worked example and every practice item is the standard sequential (chain-substitution) bridge, computed in this fixed order:

• Volume variance = (V₁ − V₀) × (P₀ − C₀), the change in units, valued at the old contribution margin per unit. This isolates "we sold a different quantity, at the margin we used to earn."
• Price variance = (P₁ − P₀) × V₁, the change in price, valued at the new volume. This isolates "the same units now sell for a different price."
• Variable-cost variance = (C₀ − C₁) × V₁, the change in unit cost, valued at the new volume. Watch the sign: a cost that fell is written as a positive number, because it helped profit.
• Fixed-cost variance = F₀ − F₁, simply the change in fixed spend. It's signed the same way, so a fixed cost that fell is a positive, favorable number.

Add the four and you get ΔProfit exactly, every time, for any numbers. That's an algebraic identity. There's no approximation or rule of thumb involved. The terms that would otherwise double-count or go missing cancel out, because volume's effect is priced at the old margin while the price and cost effects are priced at the new volume. Prove it to yourself once with small numbers. A candidate who has checked the identity by hand never again fears being asked "does that reconcile?"

Worked example: a profit bridge in dollars

Take a single-product illustration (synthetic, US $). A company budgeted 1,000 units at a $40 price and a $25 variable cost, which is a $15 margin, against $8,000 of fixed cost. That gives a budgeted profit of 1,000 × 15 − 8,000 = $7,000. It actually sold 1,200 units, but at a $38 price and a $26 variable cost, against $8,500 of fixed cost. Actual profit came out at 1,200 × (38 − 26) − 8,500 = 1,200 × 12 − 8,500 = 14,400 − 8,500 = $5,900. So profit fell $1,100 while volume rose 20%. Once you've bridged it, that pattern should never surprise you again.

Now bridge it. Volume variance = (1,200 − 1,000) × 15 = +$3,000 is favorable, since you sold more units at the old margin. Price variance = (38 − 40) × 1,200 = −$2,400 is unfavorable, because a $2 price cut landed on all 1,200 units. Variable-cost variance = (25 − 26) × 1,200 = −$1,200 is unfavorable too, with cost creeping up $1/unit. Fixed-cost variance = 8,000 − 8,500 = −$500 is unfavorable as well. Add them up, 3,000 − 2,400 − 1,200 − 500 = −$1,100, and you land on the profit change to the dollar.

So the bridge tells a story the headline number can't. Volume growth was real and did help, but a price cut and a cost creep, reinforced by fixed-cost overrun, more than ate it. That's a completely different story from "profit fell despite more sales", and a far more actionable one. It took four short multiplications to prove.

{
  "skill": "Drafting a first-pass exhibit table or bridge scaffold quickly, without letting the tool own the arithmetic",
  "use": "Ask a chat assistant to lay out a clean bridge table (volume / price / variable-cost / fixed-cost rows, with the sequential formulas in each cell) for a case you are structuring, so you spend your limited time on the numbers and the story instead of formatting — then you compute every cell yourself and check that the column sums to the given profit change.",
  "tools": ["Chat assistants — Claude, ChatGPT", "Spreadsheet tools — for building and re-running the bridge once the structure is drafted"],
  "prompt": "I'm building a profit-change bridge with these four exhibits: <paste the volume/price/cost/fixed-cost figures for both periods>. Lay out a table with rows for Volume variance, Price variance, Variable-cost variance, and Fixed-cost variance, each with the correct formula using the sequential (old-margin-for-volume, new-volume-for-price-and-cost) convention. Leave every value cell blank — do not compute the numbers for me.",
  "verify": "Compute every cell yourself by hand or in your own spreadsheet, then confirm the four values sum to exactly the stated profit change; if they do not, the fault is almost always a sign error or an inconsistent choice of which period's volume or margin a variance uses — recheck the convention, never adjust a number to force a match (Guardrail #3, Recompute, applies to your own case math exactly as it applies to an AI's).",
  "diy": "You must be able to derive and defend the sequential bridge formulas from the plain identity Profit = V(P−C) − F without help — that derivation is exactly what this section teaches, and a partner will ask you to defend the convention, not just recite it.",
  "tools_market": "US"
}

When there is more than one product: splitting volume into quantity and mix

Real businesses rarely sell one product to one segment. Once you have two or more, the "volume variance" above hides a second story worth pulling out on its own. Did total volume grow? That's a quantity or yield effect. Did the mix between low- and high-margin products or segments shift? That's a mix effect. Most often it's some of both. Ignoring mix is one of the commonest ways an otherwise-correct bridge still misses the point. A company can grow total units handsomely while selling proportionally more of its worst-margin product, and an aggregate "volume variance: favorable" line won't tell you that.

Splitting takes only one more piece of arithmetic. First compute the weighted-average budgeted margin across all products, using budgeted volumes as weights: m̄₀ = Σ(V₀ᵢ × m₀ᵢ) / ΣV₀ᵢ. Then:

• Quantity variance = (ΣV₁ − ΣV₀) × m̄₀, the change in total units, valued at the blended old margin. This answers "ignoring any mix shift, did we simply sell more or fewer units?"
• Mix variance = Σᵢ [(V₁ᵢ − ΣV₁ × (V₀ᵢ / ΣV₀)) × m₀ᵢ]. For each product, compare its actual volume to "what it would have sold at the budgeted mix percentage applied to the new total volume," and value that gap at that product's own budgeted margin. This isolates the pure effect of the mix shifting, holding total volume fixed at its actual level.

These two always sum exactly to the single-product volume variance you already know how to compute, Σᵢ (V₁ᵢ − V₀ᵢ) × m₀ᵢ. So nothing about the four-line bridge above changes. You're just opening the "volume" line into two labeled sub-lines when more than one product is in play.

Worked example: quantity and mix in rupees

Here's a worked illustration (synthetic, India ₹). A packaged-snacks maker budgeted equal volumes of two lines, Namkeen (savory) and Sweets, at 1,00,000 units each. Namkeen is priced at ₹50 with ₹30 variable cost, a ₹20 margin. Sweets sells at ₹80 with ₹50 variable cost, a ₹30 margin. Because the budgeted split is 50/50, the blended budgeted margin is simply the average, m̄₀ = ₹25. Actual volumes came in at 90,000 units of Namkeen and 1,30,000 of Sweets. Total volume rose from 2,00,000 to 2,20,000, a real 10% increase, and the mix shifted hard toward the higher-margin Sweets line.

Start with quantity. Quantity variance = (2,20,000 − 2,00,000) × 25 = ₹5,00,000, favorable, because that's real growth. Now mix. At the budgeted 50/50 mix, 2,20,000 total units "should" have split 1,10,000/1,10,000. Namkeen actually came in 20,000 below that expectation, and at its own ₹20 margin that's −20,000 × 20 = −₹4,00,000. Sweets came in 20,000 above it, and at its ₹30 margin that's +20,000 × 30 = +₹6,00,000. Together they give a mix variance of −4,00,000 + 6,00,000 = +₹2,00,000. That's favorable too, because the shift moved volume toward the better-margin product.

Now check it. Quantity of ₹5,00,000 plus mix of ₹2,00,000 equals ₹7,00,000. That must equal the plain single-product volume variance, and it does: Σ(V₁ᵢ−V₀ᵢ)×m₀ᵢ = (90,000−1,00,000)(20) + (1,30,000−1,00,000)(30) = −2,00,000 + 9,00,000 = ₹7,00,000. Both real growth and a favorable mix shift were pulling profit up here. Worked example 2 shows the opposite pattern, and it's far more dangerous: total volume grows while the mix shifts the wrong way.