The Program That Was Quietly Losing Money for Eight Months

Farhan runs a precision-turning shop on the outskirts of Pune — around ₹40 crore in turnover, a good spread of CNC turning centres, the kind of Tier-2 supplier that feeds components into the larger Tier-1s who feed the carmakers. He’s been at it twenty-odd years. He knows his machines, he knows his people, and he’d tell you he runs a tight operation. By most measures, he does.

I’m telling his story because of one program that looked, by every signal he could see, like one of his good ones — and was quietly losing him money for eight months before anyone realised.

The situation

A couple of years back, Farhan won a nice program — a family of turned components for a steering assembly, decent volume, a customer he liked. He’d quoted it carefully, the way he quotes everything: material, cycle time, scrap, overhead, a fair margin. The nomination came through, the line got set up, and the parts started flowing. From that point on, in Farhan’s mind and in everyone else’s, it was a won program doing what won programs do. The machines ran, the parts shipped, the invoices went out. Nobody had any reason to look at it again. Why would they? It was working.

What nobody saw — because nobody could see it in one place — was that the margin on that program had started slipping almost from the beginning. Three small things, each invisible on its own. The bar stock he was buying had crept up in price over the months, the way steel does, and nobody had connected that rising input cost back to this specific part’s profitability. The actual cycle time on the job was running a few seconds longer than the quote had assumed — a tiny gap per part, but this was a high-volume program, hundreds of thousands of pieces. And scrap on one tricky operation was running a couple of points higher than planned, quietly turning good material into rejected parts.

Not one of those three was alarming. Steel prices rise for everyone. A few seconds is nothing. Two points of scrap is within the range you shrug at. But stack them together, on a high-volume program, across eight months, and they had quietly turned a healthy-margin job into one that was barely breaking even — and on the worst months, losing money on every part it shipped.

Why it stayed invisible

Here’s the honest part, and it’s not a knock on Farhan. The information that would have revealed this was all there — but it was scattered exactly the way it always is. The rising bar-stock price sat in his accounts in Tally. The real cycle times sat on the shop floor, in the machines. The scrap sat in quality records. The original quoted cost sat in a spreadsheet that had been filed away the day the program was won. To see the problem, someone would have had to pull all four together, for this specific part, and compare what it was actually costing now against what it was quoted at — and then do that regularly, for every program in the shop.

Nobody does that. Nobody can. Farhan has dozens of active part numbers. Reconstructing the true, current cost of every one of them by hand, every month, would be a full-time job that produces numbers already out of date by the time they’re finished. So the program ran on its original assumption — “this is a good-margin job” — long after that assumption had stopped being true. The quote said it was profitable. Reality disagreed. And there was no single place where the two were ever put side by side.

What changed

When Farhan first got a unified view of his business — every program’s quoted cost sitting next to its real, current cost, drawn live from the material, machine, and scrap data — this program stood out immediately. Not because anyone suspected it, but because the gap between what it was quoted at and what it was actually costing had grown wide enough to be impossible to miss once the two numbers finally sat together.

The figure that mattered: across those eight months, the erosion on that one program had quietly cost him a little over ₹18 lakhs in margin he’d assumed he was earning and wasn’t. Not stolen, not through any disaster — just slipped away, a few paise per part at a time, on a job everyone had stopped watching because it looked fine.

What he did with it

Again, the fix was unremarkable, which is the point. Seeing the three causes laid out, Farhan did three ordinary things. He went back to his bar-stock supplier and renegotiated, having now connected the rising input price to a specific program’s profitability for the first time. He had his team look at the operation that was scrapping high, and a modest tooling and process change brought it back into line. And the cycle-time gap, once visible, prompted a small adjustment that recovered most of the lost seconds. None of it was clever engineering. It was the kind of thing Farhan is perfectly capable of doing in his sleep — he simply hadn’t known there was anything to do, because the problem had never surfaced.

Within a couple of months, the program was back to the margin he’d originally quoted. More importantly, he could now see the same comparison for every other program in the shop — and found two more that were drifting, caught early this time, before they’d done eight months of quiet damage.

The takeaway

Farhan didn’t lose that ₹18 lakhs to anything dramatic. He lost it to a program that looked fine — and “looks fine” is exactly the kind of program that bleeds, because nobody watches the jobs that aren’t obviously in trouble. The margin was eroding in plain sight the whole time, recorded across his own systems, in numbers he owned. The only thing missing was the one view that put quoted cost and real cost side by side, early enough to act.

That’s what being warned before the disaster actually means in an auto-parts shop. Not catching a catastrophe — catching the quiet, ordinary erosion on a won program while there are still years of production left to protect. In a business where the price is fixed for the life of the program, knowing your real cost while you can still do something about it isn’t a nicety. It’s the whole difference between a program that makes money and one that only looks like it does.

Part of the Auto-Components series. This is the central question of You Won the Part. Are You Sure You’re Making Money on It? shown in practice — and the auto-parts cousin of How a Mumbai Manufacturer Caught ₹75 Lakhs Before It Slipped Away.