Cost-Reduction Cases That Don't Collapse Into a List
Many candidates answer a cost-reduction case by listing cost categories. A stronger answer finds where the money is, picks a lever per cost line, and nets the savings against one-time cost and side effects. Worked example included.
Ask a candidate to cut costs at a company and you will often hear the same thing: "I'd look at labor, materials, rent, marketing, technology..." It sounds structured. It is a list. Neither the interviewer nor the candidate can tell which items matter, so the next ten minutes drift through categories and end with "I'd look at all of them."
A better answer has an order. Confirm the target and the constraints. Break cost down so the structure follows how money is actually spent. Find where it concentrates. Compare with a baseline. Choose levers line by line. Then net the savings against what they cost and what they might break. This post walks through that order with a worked example whose numbers add up.
Step 0: Confirm that cost is the problem
Even when the prompt says "reduce costs", spend thirty seconds on three clarifications. What is the target, and by when? What is the baseline: which costs, and which year? And what is off the table: layoffs, quality floors, service commitments?
Then ask whether cost is really what moved. If margin fell because price dropped or the mix shifted toward lower-margin products, cutting cost treats a symptom. The profitability framework is how you check: price, volume, variable cost, fixed cost, with mix layered on top. Sometimes the right answer to a cost case is "cost is part of it, but pricing is bigger."
Step 1: Build a cost tree, not a category list
The tree below follows how costs behave, which matters because different costs respond to different levers.
- Total cost = variable cost + fixed cost
- Variable cost = volume × unit cost, where unit cost is materials per unit + labor per unit + freight per unit (and any other cost that scales with output)
- Fixed cost = plant overhead, administration, R&D and anything else that does not move when volume does
The split tells you where to look. Volume-driven costs respond to unit-cost levers such as price, spec and process. Fixed costs respond to scope, footprint and headcount decisions. A single "reduce costs" lever for both is a sign the structure is too coarse.
Step 2: Find where the money is
Ask for the cost breakdown and sort it by size. A lever on a line that is 3% of cost cannot deliver a 5% target, however clever it is. Here is the example we will use. The numbers are fictional and round on purpose.
Calder Appliances makes kitchen appliances. Revenue is $900M and total operating cost is $800M, so operating profit is $100M, an 11.1% margin. Management wants $40M taken out of the annual cost run-rate, which is 5% of the cost base. Where would you look, and what would you recommend?
| Cost line | $M | Share of total | Behavior |
|---|---|---|---|
| Materials and components | 320 | 40% | Variable |
| Factory labor | 160 | 20% | Partly variable (overtime flexes) |
| Freight and logistics | 80 | 10% | Variable |
| Plant overhead and energy | 80 | 10% | Mostly fixed |
| Sales, general and administrative (SG&A) | 120 | 15% | Mostly fixed |
| R&D and other | 40 | 5% | Fixed |
| Total | 800 | 100% |
Out loud: "Three lines (materials at 320, factory labor at 160 and SG&A at 120) add up to $600M, or 75% of the cost base. A $40M target is 5% of everything, so I'll start with those three."
Step 3: Compare with a baseline
"Labor looks high" means nothing without a reference. Compare each large line against three baselines where you can: the company's own history on a per-unit basis, a competitor or industry benchmark, and the best-performing site or team inside the company. The gap to the best internal performer is often the most credible estimate of what is achievable, because the company has already shown it can be done.
Step 4: Choose levers line by line
For each large line, run five levers and pick the one or two that fit:
- Eliminate. Stop doing it at all.
- Reduce. Do less of it, or do it more efficiently.
- Substitute. Replace it with a cheaper alternative that does the same job.
- Renegotiate. Pay less for the same thing.
- Automate. Replace manual effort with a system or machine.
Before you pick a lever, go one level down in the biggest line. Materials cost is quantity per unit times price per unit, so ask how it splits: commodity inputs versus specialty parts, contracted versus spot purchases, single-source versus multi-source. Those splits decide which lever is plausible. You cannot renegotiate your way out of a part only one supplier makes, and you would not redesign a commodity input.
For Calder, you might work it like this. Of the $320M in materials, $200M is commodity inputs on contracts coming up for renewal, so renegotiate. A different $100M is specialty parts where an alternative design performs similarly, so substitute. The last $20M is small and single-source, so leave it. Freight at $80M can be reduced by shipping fewer, fuller loads. In factory labor, automating one packaging line cuts overtime and manual handling. In SG&A, stop low-value reports and duplicate regional roles.
| Lever | Cost line and assumption | Run-rate saving | One-time cost |
|---|---|---|---|
| Renegotiate | Materials: 5% lower prices on $200M of commodity spend | 10 | 1 |
| Substitute | Materials: 8% saving on $100M of specialty parts | 8 | 6 |
| Reduce | Freight: 10% lower on $80M from fuller shipments | 8 | 2 |
| Automate | Factory labor: 5% on $160M from one automated line | 8 | 16 |
| Eliminate | SG&A: 5% on $120M from dropped reports and duplicate roles | 6 | 4 |
| Total | 40 | 29 |
Check the arithmetic the way you would aloud. Savings: 10 + 8 + 8 + 8 + 6 = 40. One-time costs: 1 + 6 + 2 + 16 + 4 = 29. The plan meets the target exactly. A plan that lands exactly on target has no buffer, so in practice you would identify 10 to 20 percent more than the target. Plant overhead and R&D stay untouched for now.
Step 5: Net the savings against the cost to get them
A run-rate saving is not free. Divide one-time cost by annual saving to get payback. For the whole plan: $29M ÷ $40M = 0.725 years, about nine months. By lever it varies a lot:
- Renegotiation: $1M ÷ $10M = 0.1 years, about a month.
- Freight: $2M ÷ $8M = 0.25 years, about three months.
- SG&A: $4M ÷ $6M, roughly eight months.
- Substitution: $6M ÷ $8M = 0.75 years, nine months.
- Automation: $16M ÷ $8M = 2 years. This is the long pole.
This is also how you separate quick wins from structural moves. Renegotiation and freight deliver $18M a year (10 + 8) for $3M (1 + 2) and can start now. Substitution, automation and SG&A deliver the other $22M (8 + 8 + 6) for $26M (6 + 16 + 4) and take longer. Sequencing matters: the early savings build credibility and can help fund the slower projects, so a good plan starts with the quick wins and runs the structural work in parallel. Savings also ramp, and $29M of one-time cost lands before most of the savings do, so year one can be worse than today. At full run-rate, with revenue unchanged, operating profit moves from $100M to $140M, and the margin from 11.1% to 15.6% ($140M ÷ $900M).
Step 6: Check what the savings break
Every lever has a side effect to name. Substituting a part can raise warranty claims. Fuller shipments can lengthen delivery times. Squeezing suppliers can strain supply continuity. Automation affects people and morale. Cutting SG&A can reduce selling capacity and with it, revenue.
The revenue effect deserves arithmetic. Suppose materials and freight ($320M + $80M = $400M) are the only costs that move with volume. Then contribution margin is ($900M − $400M) ÷ $900M, about 55.6%. If a $6M SG&A saving costs you sales, it is wiped out by a revenue loss of $6M ÷ 0.556, about $10.8M, which is only about 1.2% of revenue. It takes only a 1.2% revenue loss to cancel that saving, which is why the check is worth making.
Interviewers listen for whether you tested your own recommendations. Saying "the risk is X, and here is the revenue that would cancel the saving" is a stronger finish than quoting a total and stopping.
How the recommendation sounds
"I'd recommend a five-part program that takes $40M out of the $800M cost base, or 5%. Two quick wins, renegotiating commodity contracts and consolidating shipments, deliver $18M a year for about $3M. Three structural moves, substituting specialty parts, automating a packaging line and trimming SG&A, add $22M for $26M. The total one-time cost is $29M, so the program pays back in roughly nine months at full run-rate savings, and later once the ramp-up is counted and lifts margin from 11.1% to 15.6%. The main risks are quality on the substituted parts and delivery times from consolidated freight, so I'd pilot both before rolling out. Next, I'd validate the supplier price gap with procurement."
Notice the shape: the answer first, the numbers second, the risks named, and a next step. If you struggle to say the payback math smoothly, the post on narrating calculations has a routine for it. For a full walkthrough of how a profitability case flows from prompt to recommendation, see the coffee chain example.
Two flavors of cost case
Listen for which one you have been given. In the first, volume is stable and the goal is a lower unit cost: the Calder example, solved with price, spec and process levers. In the second, volume has fallen and the fixed base no longer fits it, so the question is how to right-size. The levers change: footprint (close or merge a site), shifts and staffing, and the scope of overhead functions. The timing question also changes. Fixed-cost moves are slower and carry larger one-time costs, such as severance or lease exits, so the payback math matters even more. State which flavor you think it is before you structure, and the rest of your answer follows.
Mistakes that collapse the case into a list
- Naming categories instead of sizing them. If you have not asked for the breakdown, you do not know where to start.
- Across-the-board cuts. "Take 5% out of every line" is easy to say, hits valuable and wasteful spend equally, and ignores that some lines cannot be cut.
- Quoting savings without costs. Without one-time cost and payback, you cannot say whether a lever is a quick win or a long project.
- Ignoring side effects. Quality, service, supply and revenue are where cost programs go wrong in practice.
- Skipping the diagnosis. Not checking whether cost is actually the problem.
The Profitability Framework Beyond Revenue minus Cost
The structure to reach for first, so you know whether a cost program is the right answer at all.
One short, useful email per week.
New case frameworks, mock interview tactics, and recruiting intel from the CaseGrade team. Built for serious candidates — no fluff, no upsell every paragraph.
Unsubscribe anytime. We never share your address.
Keep reading
All articles →The Week Before Your First-Round Interviews: A 7-Day Checklist
The last week before a first round is for consolidating, not cramming. A day-by-day plan covering your final mock, logistics, firm refresh, interview-day routine and the notes to take afterwards.
Clarifying Questions: What to Ask, What Not to, and When to Stop
Clarifying questions are the first thing an interviewer hears you think. Learn the four jobs they do, the test that separates a good question from filler, what not to ask, and when to stop and start structuring.
Delivering a 90-Second Fit Story Out Loud
A story that reads well can still fall apart when you say it. Here is a five-beat structure with a word budget, a fully worked example, and the drills that turn a good story into a clean ninety-second answer.