Your company sells two products. The P&L says both are profitable at 22% margin. The commercial director recommends raising premium product volume because "it has better mix." Three years later, that strategy destroyed value — the premium was never truly profitable. The standard product was subsidizing the premium, without anyone noticing, because traditional costing allocated overhead by volume rather than real complexity.
Activity-based costing (ABC) is the method that reveals that hidden truth. Instead of allocating overhead "by volume" (which favors simple, large-scale products), ABC measures which activities each product actually consumes — setup time, customer service, quality control, engineering, support — and charges cost proportionally. The company that knows its ABC knows which products genuinely make money and which are subsidized; without ABC, pricing and mix decisions are made on numbers that structurally deceive.
In this module you learn the mechanics of ABC, when it's actually worth applying, and the three errors that kill implementations (overengineering, lack of operational buy-in, and "completeness" instead of selectivity).
Let's go.
Your company manufactures two products. Product A sells 1,000 units/year at $800 each. Product B sells 8,000 units/year at $500 each. Total overhead is $400K. Under traditional costing (allocated by volume), who carries more overhead? Is that correct?
In plain language
Before the mechanics, the four basic questions.
Why do this at all?
Because traditional costing systematically deceives in companies with complex mix. High-volume / low-complexity products end up subsidizing low-volume / high-complexity products. The company believes its premium mix is profitable when it's actually loss-making. Without ABC, pricing, mix, and portfolio decisions are made on fictitious data. With ABC, operational truth is revealed — and often very different from what the traditional P&L suggested.
Who builds it and who uses it?
CFO + senior FP&A design the model and identify key activities. Operations / Supply / Customer Service provide real-time consumption data per product. CEO and executive committee USE results for pricing, mix, portfolio decisions. Without operations participation, the model is theoretical and not believed. Without executive committee use, the exercise is accounting theater.
When does it show up?
Typically every 3-5 years, not annually. Deep ABC is done when: (a) the company has a complex and heterogeneous product mix. (b) there's suspicion of cross-subsidies (a supposedly profitable "premium" that nonetheless doesn't scale as expected). (c) a big strategic decision is coming (launch, discontinuation, material repricing). In between, traditional costing keeps running for day-to-day reporting.
What if we never do it?
The company operates on false assumptions about per-product profitability. Subsidizes unprofitable products without knowing. Discontinues profitable products because they "have low margin" when calculated margin is distorted. Raises volume of products that destroy per-unit value. A company with complex mix without ABC for 5+ years typically erodes 5-15% of potential EBITDA — invisible until a consultant or buyer discovers it in due diligence.
Andina S.A. — discovering the cross-subsidy
Andina launches a new premium line 3 years ago. Traditional P&L shows the line generates 28% margin vs 22% for the standard line. Commercial director pushes for more premium volume. Three years later, total EBITDA dropped 4pp despite premium mix growing from 15% to 35% of revenue.
CFO commissions ABC analysis. Result: the premium line does not generate 28% margin — it generates 8%. Reason: traditional costing allocated overhead (plant, customer service, engineering, quality) proportional to volume. Premium represented 15% of units sold but consumed 45% of setup hours, 60% of customer service time (more sales complexity), and 70% of quality checks (higher standards). Real overhead per premium unit was 3.2x that of standard.
Under corrected ABC costing: premium 8% margin, standard 27% margin. The 3-year strategy (prioritize premium) destroyed value. Each incremental premium unit replaced a standard unit that was 19pp more profitable — multiplied by thousands of units, $4-6M of EBITDA per year lost.
The board conversation changes. They don't discuss "should we make more premium?" but "does premium have a path to 25%+ margin via complexity reduction, or do we partially discontinue?". Without ABC, that conversation is never had — the company keeps pushing for more premium believing it creates value.
The visual below lets you play with two hypothetical products and see how the costing method changes the verdict on which actually makes money.
Traditional vs ABC, live
Two products: Premium Service (high complexity, low volume) and Standard Service (low complexity, high volume). Move the activity sliders that each consumes — setup, customer service, quality control. Watch the gap between traditional and ABC margin.
The critical experiment: raise Premium setup hours to 3.0h and lower Standard to 0.2h. Premium goes from "profitable under traditional" to "loss-making under ABC." That's the magnitude of the cross-subsidy — and the reason ABC matters.
Interactive visual
Traditional costing vs ABC — two products, two truths
Two products that look equally profitable under traditional costing. Move each one's operational complexity and watch ABC reveal the hidden truth: one makes money, the other loses it.
Premium Service (low volume)
Activity consumption per unit
Setup / preparation (hours)
2.5h
Customer service (hours)
1.5h
Quality checks (inspections)
3.0
Standard Service (high volume)
Activity consumption per unit
Setup / preparation (hours)
0.3h
Customer service (hours)
0.2h
Quality checks (inspections)
0.5
Annual margin: traditional vs ABC
What you are seeing
Three critical lessons: (1) Traditional costing allocates overhead proportional to volume — Standard Service pays most because it sells more units. But if Standard is genuinely simple and Premium is complex, Standard is SUBSIDIZING Premium. ABC corrects that: each product pays for the activities it really consumes. (2) The gap between traditional and ABC margin is the magnitude of the hidden cross-subsidy. When that gap is >5pp of price, pricing and mix decisions made with traditional costing are systematically wrong. (3) ABC is not magic — the data is the same. What changes is the allocation method. Total company costs are identical. What changes is WHICH product carries them. That redistribution sometimes turns a "winner" into a "hidden loser" and vice versa.
The critical reading of the visual: the GAP between traditional and ABC margin is not a calculation error — it's the magnitude of the lie you were operating on. If Premium has ABC margin 8% and traditional margin 28%, that 20pp gap is the amount by which Standard subsidizes Premium per unit. Multiplied by annual volume, it's millions of dollars misallocated.
Second reading: ABC is not magic. Total costs are the same. What changes is WHICH product carries them. If Premium really consumes 45% of overhead and only pays 15% under traditional, Standard pays the rest without anyone thanking it. ABC simply aligns cost with real activity consumption.
Third reading: ABC should not be applied to ALL products or ALL costs. Selectivity is the discipline. Apply ABC to the 5-10 most representative SKUs of your mix, to the most material overhead costs (>$500K), and to decisions where 5pp of margin matters. Applying it "to everything" generates 12-month implementation theater that no one uses afterward.
The mechanics: how to implement ABC without killing the team
- Identify the 5-10 activities consuming >70% of your overhead. Setup, customer service, quality control, engineering support, distribution, IT, etc. Each activity needs its own "rate" (cost/hour or cost/transaction). Do NOT try to model every activity — only the ones that move the number.
- Measure activity consumption per SKU via time studies, not estimates. One month of tracking: how many setup hours per product, how many customer service calls per unit, how many quality checks. Without real data, ABC becomes glorified judgment.
- Prioritize material SKUs — not all of them. Pareto: 20% of SKUs typically represent 80% of revenue AND 80% of cost. Apply ABC to those 20%. For the rest, traditional costing with reasonable mark-up is enough.
- Compare ABC margin vs traditional. Document material gaps (>5pp). Those gaps are where you make wrong decisions. Each SKU with gap >5pp deserves specific review of pricing, mix, or discontinuation.
- Build buy-in with operations BEFORE presenting. The operations director must accept that activity rates are fair. Without their buy-in, the entire model is challenged in every conversation. Unilateral CFO imposing ABC = guaranteed failure.
- Refresh the model every 18-24 months, not annually. Activity rates and assumptions change with operations. Annual refresh burns the team; refresh every 2 years maintains relevance without saturation.
- Use ABC outputs for specific decisions, not "general reporting". Quarterly pricing review. Decision to discontinue SKU. New product launch. Supplier renegotiation. When ABC informs tangible decisions, the exercise justifies itself. When it only appears in reports no one uses, it's theater.
- Cost driver. The unit measuring activity consumption. Setup time = hours. Customer service = calls/hours. Quality = inspections. The driver must correlate strongly with the activity cost.
- Activity rate. Total activity cost divided by cost driver units. E.g.: $100K total QC / 4,000 inspections/year = $25/inspection.
- Resource consumption. How much of the cost driver each product consumes. Premium product = 3 inspections/unit. Standard = 0.5/unit.
- Allocated activity cost. Activity rate × resource consumption. Premium QC cost = $25 × 3 = $75/unit. Standard QC cost = $25 × 0.5 = $12.50/unit.
- Traditional vs ABC. Traditional allocates overhead by a simple metric (volume, machine-hours). ABC allocates by real consumption of each activity.
- Time-driven ABC (TDABC). Modern variant using "minutes per transaction" instead of complex activity tables. Simpler to maintain.
- Practical rule: if your mix is relatively homogeneous (all products similar in complexity), ABC adds little value. If your mix is heterogeneous (premium + standard + custom + commodity), ABC almost always reveals material cross-subsidies.
Adversarial check
Adversarial check
1.Your commercial director pushes for more premium volume arguing "it has 28% margin vs 22% standard." What do you ask BEFORE approving?
2.Why is applying ABC to ALL SKUs instead of selected ones the most common error?
3.Your ABC analysis reveals a "premium" SKU has real margin of 8%, not the reported 28%. What is the right response?
Exit checklist
Suggested re-review: every 18-24 months with selective ABC analysis of the 20 material SKUs. And whenever a big pricing, mix, launch, or discontinuation decision appears — those are the moments where ABC informs tangible decisions.
Optional
Go deeper
Sources and books to dig into the original material