Why Your Capacity Forecast Is Wrong (And Why It’s Not Vardhman's Fault)

The Forecast That Almost Sank a Q3 Launch

In April 2023, I submitted a capacity projection for our upcoming fall line. I'd spent two weeks on it. I'd cross-referenced Vardhman's public announcements, talked to our account manager, and even factored in the monsoon season.

I was so confident, I told my boss, 'We're solid. We've got the yarn.'

I was wrong. Not about Vardhman's ability to produce, but about how I was interpreting their capacity. The forecast missed by 35%. That error cascaded—late sample approvals, missed production windows, and a scramble for secondary suppliers that cost us roughly $12,000 in premium pricing and expedited freight.

It wasn't Vardhman's fault. It was mine. I'd made the classic mistake of treating a capacity figure like a hard limit, when it's actually a dynamic variable. Let me show you what I mean.

The Surface Problem: Nobody Trusts the Numbers

When I talk to other procurement folks, the complaint is almost always the same: 'I can't trust the capacity forecasts from big mills.' They point to a project from Q1 2024 where a supplier like Vardhman quoted a 4-week lead time, and it stretched to 6. They look at a public capacity figure of, say, 10 million kg per month, and their actual allocation feels like a fraction of that.

The assumption is that the supplier is hiding something. That there's a secret reserve of capacity they're holding back for bigger clients. People think the problem is a lack of transparency from the supplier.

But here's the thing you learn after a few years in this business: the supplier's numbers aren't the problem. The problem is the assumption that 'capacity' is a single, static number.

The Real Reason: You're Reading the Wrong Map

I've seen this pattern play out more times than I can count. On a 50,000-meter order for a denim jacket program, our team expected a seamless run. The mill had the capacity, we'd done the math. But we didn't ask the right questions.

What most people don't realize is that a published 'capacity' figure is usually a theoretical maximum under ideal conditions. It assumes a perfect mix of products, no material shortages, full labor availability, and zero machine downtime. That's not real life. Real life is a maddening cascade of trade-offs.

Here's something vendors won't tell you in a first meeting: every new order is a negotiation for a slice of a very complicated pie. A large, diversified company like Vardhman isn't just spinning yarn. They're juggling cotton, wool, and acrylic programs. They're managing different counts, different twists, different dye lots for different clients.

The assumption is that capacity is about the machines. I used to think that too. But the real bottleneck isn't the spinning frame. It's the setup time, the color change, the yarn-dyeing schedule, and the availability of specific cotton grades. A 10-day run of a single, simple yarn is wildly different from a 10-day run of 10 different specialty blends. The capacity feels different even though the machine hours are the same.

Let me give you a concrete example. In August 2022, I needed 20,000 kg of a specific cotton-plus yarn. Vardhman's system said they had capacity. My account manager confirmed it. It looked fine on the screen.

The reality was that the same machines were scheduled for a massive woolen program for another client. Our cotton yarn required a full cleaning cycle and a change in drafting settings. It's a 6-hour setup. Do that across multiple machines, and suddenly your 'available capacity' is chewed up by transitions.

The Price of False Assumptions

I made that mistake—treating the capacity forecast as a guarantee—on a $3,200 order for a specialty yarn. I'd promised my internal stakeholder a 4-week delivery. When the mill pushed back to 6 weeks, it wasn't a disaster for the order itself, but it was a disaster for my credibility.

That error cost me a 1-week delay and $890 in premium freight to get the material to our cut-and-sew factory on time. Worse, the production planner started to treat all my forecasts with skepticism. A reputation hit that took three months to repair.

We've caught 47 potential errors using this checklist in the past 18 months.

Wait, let me rephrase that. I don't have a magic checklist yet—that's the system we're building now. But I do have a mental model that's saved us a ton of time and money.

Think about the real costs of a bad forecast:

• Missed revenue: Your product launches late.
• Emergency costs: You pay 25–50% more for rush orders from smaller, less reliable mills.
• Internal friction: Sales blames you, you blame the mill, the mill points at your specs. A giant blame game.
• Lost trust: Your internal teams start building their own backup plans, fragmenting your buying power.

That third rejection in Q1 2024—a client's order being delayed because our yarn wasn't available—was the final straw. I realized we weren't just failing to predict the future. We were failing to ask the right questions about the present.

So, What Actually Works?

Once you understand that capacity is a fluid concept, the fix becomes obvious. You stop looking for a single number and start looking for a dialogue.

First, stop asking 'How much capacity do you have?' That's a trap question. It invites a theoretical answer. Instead, start asking situational questions.

• Instead of: 'What's your monthly capacity for cotton yarn?'
  Ask: 'For a repeat order of 15,000 kg of our standard 30/s cotton yarn, with a lead time of 4 weeks, what's the realistic throughput assuming current schedules?'
• Instead of: 'Can you handle a rush order?'
  Ask: 'If we shift this order to a lower-priority program, what's the first thing that gets delayed? And what would that cost us?'

Second, I've started building a simple 'capacity risk profile' for each major program. It's not a complex spreadsheet. It's a set of three questions:

1. Complexity: How many raw material changes or setup changes does this order require? (More changes = less predictable capacity.)
2. Timeline overlap: What else is running on similar machines during our window? (Ask your rep—they know.)
3. Criticality: If this slips by two weeks, what's the dollar impact?

I've started using this mental model before every significant order. In Q4 2024, it flagged a potential risk with a Vardhman order for a multicolored knit program. The setup complexity was high, and the timeline overlapped with a peak season for their wool division. The forecast was optimistic. We built in a 2-week buffer.

Did the order ship late? No. It shipped exactly on our new timeline. We looked like heroes. But more importantly, we didn't waste $12,000 learning a lesson we'd already learned.

P.S. — Prices referenced for rush shipping are based on industry averages for expedited freight as of January 2025. Verify current rates before planning. The real lesson here isn't about a single supplier's performance. It's about changing how you think about the word 'forecast.'