Your CEO opens the Q3 report. Revenue +5% over plan. She turns to you and smiles: "Great quarter."
Here's the question that defines your finance career: did your team actually beat the plan, or did you just get lucky?
Variance analysis is how you answer that question — with discipline, with clean numbers, and with the confidence of someone who knows exactly what happened. It's the technique that separates the finance professionals who explain results from those who just report them.
By the end of this module you'll have it cold. Let's walk through it together.
Plan was $80M. You closed at $74M. Of the $6M gap, how much comes from price and how much from volume?
The plain-English version
Before any formula, let's make sure the four basic questions are clear.
Why do we do this?
To know if a result was earned or accidental. Anyone can read the final number; CFOs and FP&A leaders explain where that number came from and what it says about the business.
Who uses it?
FP&A teams for monthly reporting, controllers for close and audit, business unit leaders defending their results, and CFOs in exec committee and at the board. If you present finance to anyone, this is for you.
When does it come up?
Every monthly close. Every quarterly board prep. Every time a result diverges from plan enough that someone raises an eyebrow. In practice: all the time.
What if we skip it?
You congratulate the team for a quarter that was actually the dollar, not their work. You lose credibility when the auditor decomposes what you couldn't. Your board starts doubting every number you put in front of them.
Meet Andina S.A.
Andina S.A. is a fictional coffee company selling across LATAM. We'll use it as the running example throughout this module so you don't have to relearn the context every time.
At the start of the fiscal year, the FP&A team built a plan: $100M in revenue. Twelve months later, the actual landed at $117M.
The gap is $17M favorable. Good news, right?
It depends. And "it depends" is exactly the problem variance analysis solves.
That $17M didn't come from one place. It came from price decisions, from volume decisions, from the mix of products customers actually bought, from the peso moving against the dollar. To say anything intelligent about whether your team earned those $17M — or just caught a tailwind on FX — you have to break the gap down into its real drivers.
That's variance analysis. The technique is just that: decomposition. You take one number (the $17M gap) and split it into a handful of named, additive components, each one explainable on its own.
The math is grade-school. The judgment for building it well — that's what you'll learn here.
See the bridge below. Step through it to watch the story unfold.
Variance bridge — step by step
Andina S.A. · FY revenue ($M)
Andina entered the FY with a plan of $100M in revenue. This is the baseline — every variance is measured against this number.
Look at what you just did. You started with a vague question ("was the result good?") and ended with a precise answer: the team added $11M of real value, $4M came from FX (not repeatable), and there's a yellow flag in the mix shift worth understanding.
That's the output a CEO or board expects from finance. Not the number — the story behind the number.
The rest of this module covers the formal techniques for doing this well: PVM, Rate-Efficiency, FX decomposition. But the heart of variance analysis is what you just did: break a gap into named parts, each one attributable to a decision or an event.
- Variance = Actual − Baseline. But always name which baseline you're using: Plan, Forecast, Prior Year, or Prior Quarter. They are not interchangeable.
- Sign convention: F (Favorable) = improves operating result; U (Unfavorable) = degrades it. Apply consistently regardless of P&L line direction.
- The trap: variance vs. Plan ≠ variance vs. Forecast ≠ variance vs. LY. Report all three when material and label which one drives the narrative.
The three core decompositions
In the Andina bridge you saw the shape. Now the three decompositions every finance function uses, in order of how often they show up. Each one earns a story and a visual — let's go.
1. Price–Volume–Mix (PVM)
Back to Andina. In the bridge you saw +$10M Volume and +$5M Price. Where do those numbers actually come from?
Say last year they sold 100,000 bags at $50 each. This year: 140,000 bags at $70 each. Revenue jumped from $5.0M to $9.8M — a $4.8M gap.
Volume? Price? Both? You can't just say 'sold 40% more' and 'priced 40% higher' and call it +80%. That double-counts. You need a discipline: decomposing the geometry.
Applies to revenue and to COGS. Volume = ΔQ × baseline P. Price = ΔP × Q (convention determines whether actual or baseline Q). Mix = the residual when composition shifts. The interaction ΔP × ΔQ must be allocated explicitly.
- US/CPG convention: Volume = ΔQ × P₀; Price = ΔP × Q₁; interaction lives inside Price.
- Textbook convention: Volume = ΔQ × P₀; Price = ΔP × Q₀; interaction is a separate line.
Drag the sliders below. Watch the rectangles re-arrange as ΔQ or ΔP changes — and notice how the Interaction region "hides" inside Price when you switch to US/CPG.
PVM rectangle explorer
Drag the sliders. Watch how Volume, Price, and Interaction change.
2. Rate–Efficiency
Andina's roastery had a rough year. The plan: 1,000 hours at $20/hr to roast the target volume — $20K total cost. Actual: 1,100 hours at $22/hr — $24,200. That's $4,200 over budget.
Where did the overage come from? The new wage agreement (rate)? The team being slower than planned (efficiency)? Or both?
The Rate–Efficiency decomposition answers this in two clean numbers.
Applies wherever you have an input × a rate: labor hours × wage, machine hours × hourly cost, fleet miles × fuel rate. The decomposition is always the same:
• Rate variance: did the price-per-unit change?
• Efficiency variance: did we use more (or fewer) units than standard?
Recommended textbook convention: Rate variance = (R₁ − R₀) × H₁ (uses actual hours); Efficiency variance = R₀ × (H₁ − H₀) (uses standard rate). The interaction (ΔH × ΔR) lives inside Rate.
For absorbed overhead, use three-way: Spending + Variable Efficiency + Volume.
Rate–Efficiency rectangle
Andina's roastery · standard 1,000 hours × $20/hr vs actual 1,100 hours × $22/hr
The Rate strip sits across the FULL actual width (1,100 hours). That's the textbook convention — Rate variance carries the interaction (ΔH × ΔR = $200) inside it. If you used standard hours instead (1,000), Rate would be $2,000 and you'd surface the $200 interaction as its own line.
3. FX decomposition
Andina has subsidiaries in Chile, Mexico, and Colombia. Each grew in local currency. But this year the dollar strengthened ~10% against every peso.
When the PE board asks 'how did the LATAM portfolio do?' there are at least three correct answers — and the wrong move is to pick one and pretend the others don't exist.
FX decomposition applies to any multi-currency operation. Four lenses you have to be able to show:
• Local: each subsidiary in its functional currency. Best for judging in-market team execution.
• Reported: consolidated USD using actual exchange rates each period. This is what hits the consolidated P&L. Use for USD-functional investors.
• Constant Currency (CC): USD using fixed exchange rates (typically prior year). Isolates operational performance from FX. Use to answer 'what's the team and what's the dollar.'
• Transactional vs Translational: separately, distinguish real cash gains/losses on FX-denominated transactions (transactional, hits cash) from non-cash impacts of consolidating subsidiary financials (translational, hits accumulated OCI).
Reporting all four when material is the discipline. Picking one and hiding the others is the trap.
FX decomposition · same business, three lenses
Andina's LATAM portfolio · YoY revenue growth. The dollar strengthened ~10% against each peso this year.
The lesson
Read the difference between Reported and Constant Currency. That gap — about 9 points here — is pure FX impact. Same business. Same units. Same prices. The dollar moved.
deabaco · deabaco.com
Section 2 — Current-state checklist
What's evolved in the last 5 years. Check each item you already handle. If you score <6 of 8, there's modern practice worth a deeper look.
Adversarial check
Adversarial check
1.Your Q3 revenue is +5% vs. plan. Volume flat. Prices +8%. Mix shifted heavily toward lower-margin SKUs. Your CEO celebrates the +5%. What do you tell the audit committee?
2.A competitor raised prices 12%. Operations wants to follow; Sales says "raise prices → volume drops 15%". CEO asks for your one-page recommendation by Friday.
3.Your LATAM subsidiary reports −10% YoY in USD. Local currency +8%. PE sponsor is USD-functional with USD-denominated return targets. How do you present?
PVM Bridge Template
PVM Bridge Template
Client-ready Excel. 12 sheets, 251 formulas, zero errors. Switchable US/CPG and textbook convention. Constant currency view. Bps GM Bridge. AI commentary prompts.
Exit checklist
Suggested re-review: 12 months. Frameworks are stable; AI tooling will evolve fastest — reset the prompts then.
Optional
Go deeper
Sources and books to dig into the original material