The Pump Maker's Vanishing Margin
The Prompt
Your client manufactures agricultural water pumps in Coimbatore, selling through 800 dealers across South India. Revenue has been steady at about ₹600 crore, but operating margin has slid from 14% to 9% over six quarters. Diagnose the cause and recommend fixes.
Structure & Hypothesis
Analysis & Data
Materials are up slightly but hedged. Labour, energy, logistics, SG&A — all flat as a % of revenue. Warranty and service costs have tripled, from ₹6 crore to ₹19 crore a year.
That's ₹13 crore of the ₹30 crore — a big piece, and tripling is a step-change, not drift. Warranty spikes usually trace to a cohort: a component change, a new supplier, or a new product line. Did anything enter the bill of materials about two years ago?
Cohort thinking — warranty cost today reflects units sold 1–2 years ago.
Yes — the client switched to a cheaper motor-winding supplier 20 months ago, saving ₹400 per pump. Field failure rates on that cohort are 3× the old supplier's. The remaining gap, by the way, is volume-driven overtime in the service network — caused by the same failures.
So the "saving" is the cause of nearly the whole margin slide — the cheaper winding saves about ₹8 crore a year on 2 lakh pumps, but generates ₹13 crore in claims plus service overtime, and is quietly damaging the brand with farmers whose pumps fail mid-season.
Recommendation
Recommend
- Revert (or dual-source) the motor winding; qualify the cheap supplier only if failure rates converge in accelerated-life testing.
- Proactively service the high-risk installed cohort before peak irrigation season — cheaper than failures plus reputation damage.
- Institute a total-cost-of-ownership gate on procurement: no component change ships without a warranty-impact projection.
- Track warranty cost per cohort, not per quarter, so the next bad batch is visible in months instead of six quarters.
Key Takeaway
What this case teaches
Cost-side cases reward a clean MECE sweep — but the elegant move is cohort logic: today's warranty line reflects yesterday's decisions. And every procurement "saving" should be priced at total cost, not unit cost.