What Downtime Really Costs an Indian Plant, and What to Measure First
SRF Capital Studio Research DeskFunding Intelligence, SRF Capital StudioGlobal downtime benchmarks, converted to rupees, produce numbers larger than most Indian plants' revenue. Here is how to calculate the real cost for your own lines, and which five numbers any analytics effort should produce first.
Summary
- The downtime figures in most software pitches come from the world's largest plants, and scaled to a ₹100 crore Indian manufacturer they overstate the loss by a factor of a hundred or more.
- The real cost of an hour lost is the contribution you could not sell, plus overtime, scrap and penalties, and it depends on whether the line that stopped is your bottleneck.
- Get five numbers right before buying any analytics platform: downtime by reason, OEE on the bottleneck, first-pass yield, repeat-failure rate and cost of poor quality.
Most of the pitches that reach Indian manufacturers for analytics or root cause software open with the same slide. Downtime costs $260,000 an hour. The average plant loses 25 hours a month. Multiply, convert to rupees, and the software pays for itself in weeks.
One estimate in circulation put the loss for a mid-sized Indian manufacturer at ₹80 lakh an hour and ₹288 crore a year. Even its arithmetic was off: 25 hours at ₹80 lakh is ₹20 crore a month, not ₹24 crore. More importantly, ₹288 crore is more than the entire annual revenue of most of the plants it was describing.
If the downtime number in the pitch is bigger than your revenue, the pitch is not about your plant.
Where the benchmarks come from
The benchmarks are real. They describe a different kind of factory. Siemens' True Cost of Downtime 2024 studied Fortune Global 500 companies and found that a large plant loses about 27 hours a month to unplanned stoppages, costing roughly $129 million a year. In automotive, an idle line in a major plant can cost up to $2.3 million an hour. Aberdeen's cross-industry average of about $260,000 an hour is drawn from a similar population.
An Indian auto-components supplier with ₹100 crore of revenue is a different animal. It might run three lines across two shifts. Its hour of downtime is worth what that hour would have earned, and that you can calculate from your own books.
Calculating your own number
Take that ₹100 crore plant, running about 6,000 productive hours a year on its main line. The steps below are illustrative, and the inputs are the ones your finance team already holds.
Worked example: the cost of unplanned downtime on one line of a ₹100 crore plant
| Step | Illustrative value | Where it comes from |
|---|---|---|
| Annual revenue through the line | ₹100 crore | Sales ledger |
| Productive hours a year | 6,000 | Shift plan less planned stops |
| Revenue per line hour | ₹1.67 lakh | Revenue divided by hours |
| Contribution margin | 30% | Price less material, power, variable labour |
| Lost contribution per hour, if sold out | ₹50,000 | Revenue per hour times margin |
| Unplanned downtime a year | 300 hours | Downtime log, 25 hours a month |
| Lost contribution a year | ₹1.5 crore | Hourly loss times hours |
₹1.5 crore a year is a serious number for a plant that size, often more than its annual IT budget. It is also about 190 times smaller than the figure in the pitch.
Two adjustments decide whether even ₹1.5 crore is right.
Is the line your bottleneck?
Lost contribution only counts if you could have sold what you did not make. If the stopped line has spare capacity, the hour is recovered on the next shift, and the real cost is the overtime, the extra power and the stress on delivery dates. If it is the bottleneck, the hour is gone for good. Most plants have one line or one machine that sets their output. Measure that one first.
What else leaks with the stoppage?
A breakdown rarely costs only the stopped hours. Restarts produce scrap. Rushed recovery produces quality escapes that come back as customer complaints. Late shipments to an OEM trigger penalties or, worse, a lower supplier rating at the next sourcing decision. Industry commentary often puts the full bill at two to three times the direct loss once these are counted, but that multiplier is a rule of thumb. Add them from your own records, not from a multiplier.
The five numbers to produce first
Manufacturing analytics, stripped of its vendor language, means turning what happens on the floor into numbers someone acts on. For a mid-sized Indian plant, five come before everything else.
- Downtime by reason code. Every stoppage longer than a set threshold, with a cause chosen from a short fixed list. Without reason codes you have a total, not a diagnosis.
- OEE on the bottleneck. Availability times performance times quality, for the machine or line that sets output. Measure one asset well before you measure all of them badly.
- First-pass yield. The share of units right the first time, by product and shift. It shows where the process is fragile.
- Repeat-failure rate. Of this month's breakdowns and defects, how many have happened before on the same asset or part? This is the number that tells you whether problem solving is working, and it is the whole case for root cause analysis.
- Cost of poor quality. Scrap, rework, returns, warranty and penalties, in rupees, each month. It turns quality from a department into a line on the P&L.
Each of these can start on a spreadsheet and a disciplined shift log. Software earns its place when the log has too many entries to read, when three teams keep three versions, or when a customer wants the trail. That threshold, and the reasons Indian plants cross it late, are the subject of why RCA software adoption is still low.
Why this matters more for Indian suppliers now
The scale of the opportunity is real, even if the pitch deck numbers are not. Manufacturing was about 17% of gross value added in 2024-25 on the old national accounts series, and 14.8% in 2025-26 on the revised series with base year 2022-23. The National Manufacturing Mission aims to take that share to 25%.
The production-linked incentive schemes have pulled in real capacity: ₹1.76 lakh crore of investment, over ₹16.5 lakh crore of sales and more than 12 lakh jobs by March 2025, according to the government, rising to about ₹2.4 lakh crore of investment by FY26. Smartphones became India's largest export item in FY25, at about $24 billion.
MSMEs sit underneath much of this. The Economic Survey 2025-26 credits them with about 35.4% of manufacturing output and 48.58% of exports, and over 6.5 crore units are registered on the Udyam and Udyam Assist portals. Many of them now supply global OEMs whose supplier quality manuals expect documented problem solving and measured performance. The buyer asks for the numbers, and the plant that can produce them keeps the business.
The customer , not the software vendor, is what will push most Indian plants to measure properly.
What to do, by role
For the MSME owner: get the downtime log and reason codes running on your bottleneck this quarter, even on paper. Put a rupee figure on it using the method above. That number decides how much any software is worth to you, and it should come before any demo.
For the CFO or finance head: own the cost model. The plant knows the hours; finance knows the margin, the penalties and the working capital tied up in rework. Report downtime and cost of poor quality monthly beside the P&L, the same way you would any other FP&A driver. This is the kind of operating number that stops being visible to a founder by instinct as a business grows, as we argue in the gut stops working past ₹20 crore.
For both: use the free options first. The Ministry of Heavy Industries' SAMARTH Udyog Bharat 4.0 programme runs smart manufacturing demonstration centres, including at CMTI Bengaluru and C4i4 Lab in Pune, where MSMEs can see sensors, dashboards and connected maintenance working before spending on them.
Once the numbers exist, they belong in the same operating rhythm as sales and cash. How production data links to the rest of a manufacturer's revenue engine is set out in RevenueOS for manufacturing, and which systems your analytics must eventually read from is covered in MES, ERP and IoT integration.
How useful was this article?
One tap. It tells us what to write more of.
About the author
SRF Capital Studio Research Desk
Funding Intelligence, SRF Capital Studio
The next one
Get what we publish next, by email.
Working notes on raising, borrowing, protecting, growing and structuring capital in India. One email a week at most, and you can leave any time.
We use your address only to send this. See our privacy policy.
We store your address to send you these emails and nothing else. See our privacy policy.
Related reading