How to Start a Small Garment Factory — The Honest Guide from Someone Who Did It

I am a medical doctor who started a garment factory. People think that is unusual. It is. But the skills that matter in garment manufacturing are not sewing — they are systems thinking, process design, and not panicking when 200 operators are waiting for fabric that has not arrived.

I did not grow up in garments. I did not inherit a factory. I walked into this industry with zero connections, zero experience, and the kind of confidence that only comes from not knowing what you do not know. Three years later, I run a factory with multiple sewing lines, built the ERP software we use on the floor, and have learned more about manufacturing than any textbook could teach.

This guide is what I wish someone had given me before I started. Not the sanitized version you find on business websites. The honest version. Including the parts where everything went wrong.

Should You Start a Garment Factory?

Before we talk about machines and layouts, let me be direct about the economics. Garment manufacturing is a low-margin, high-volume business. If you are looking for a business where you invest money and watch it grow passively, this is not it. You will be on the factory floor. You will deal with labor issues, quality rejections, payment disputes, and buyers who change specifications after you have already cut the fabric.

Here is the honest financial picture for a small factory in Nepal (30-50 operators):

The capital breaks down roughly like this: NPR 10-15 lakh for machines (if buying used), 5-8 lakh for factory setup (tables, racks, electrical, lighting, ventilation), 3-5 lakh for initial thread, trim, and consumables, and 10-15 lakh as working capital to cover 2-3 months of operator wages and rent before revenue starts flowing.

Your monthly burn rate will be NPR 8-12 lakh for a 40-operator factory: roughly 5-7 lakh for operator wages, 1-2 lakh for rent, and 1-2 lakh for electricity, maintenance, and overheads. In USD, you are looking at approximately $25,000-40,000 to start and $6,000-9,000 per month to run.

CMT margins are thin — 10-15% on a good day. That means on an order worth NPR 10 lakh, your profit before tax is NPR 1-1.5 lakh. You need consistent orders running back-to-back to make the math work. One month without orders and your working capital evaporates into wages for idle operators.

CMT vs FOB vs Jobwork — Pick Your Model

Before you buy a single machine, you need to decide how you will operate. There are three basic models in garment manufacturing, and they have very different capital requirements, risk profiles, and margins.

CMT (Cut-Make-Trim)

The buyer provides the fabric and trims. You cut, sew, and finish. You get paid a making charge per piece. This is the lowest-investment, lowest-risk model and it is where almost every new factory should start.

Why CMT is best for beginners: You do not need to source fabric (which requires supplier relationships you do not have yet). You do not carry fabric inventory risk (if a style gets cancelled, it is the buyer's fabric, not yours). Your capital is tied up only in machines, labor, and overhead — not in $50,000 of fabric sitting in your store.

The trade-off: CMT margins are the thinnest in the industry. You are selling labor, and labor is a commodity. The buyer controls the fabric, the design, and the timeline. You are essentially a service provider.

FOB (Free on Board)

You source the fabric, buy the trims, manufacture the garment, and ship it to the buyer. You handle everything from raw material to finished product. The buyer pays a single price per piece that includes fabric, making, and your margin.

Why FOB pays more: Because you are taking more risk. You source fabric (and bear the risk of fabric defects, delays, and price fluctuations). You carry inventory. You manage the entire supply chain. For that extra risk, you earn a higher margin — typically 20-30% instead of 10-15%.

Why you should not start with FOB: Fabric sourcing requires relationships with mills and fabric traders that take years to build. One bad fabric purchase — wrong shade, wrong GSM, late delivery — can blow up an entire order. I have seen experienced factories lose money on FOB orders because of a single fabric sourcing mistake. As a newcomer, you do not have the experience to manage this risk yet.

Jobwork (Subcontracting)

You take on part of the process for another factory. Maybe they are overloaded and need someone to handle their overlock operations. Maybe they need finishing done externally. You are essentially a subcontractor to another manufacturer.

Good for: Getting started with minimal orders of your own. Learning the operations. Building a team. Earning revenue while you look for direct buyers.

The problem: You have no direct relationship with the end buyer. You are dependent on the factory that gives you work. They can pull the work anytime. Margins are even thinner than CMT because you are the subcontractor, not the prime contractor.

My recommendation: start with CMT. Build your team, learn your operations, establish quality consistency, and develop buyer relationships. Move to FOB after 12-18 months when you understand fabric sourcing and have the working capital to support it.

What You Need Before Day One

Here is the checklist of everything you need to have in place before your first operator sits down at a machine:

1. Confirmed orders

I said it before and I will say it again. Do not set up a factory without at least one confirmed order with a purchase order or at minimum a written commitment from a buyer. "I will give you orders" said over tea is not a confirmed order. A purchase order with style, quantity, delivery date, and price — that is a confirmed order.

2. Factory space

For 30-50 operators, you need 3,000-5,000 square feet of usable floor space. This needs to include cutting area, sewing area, finishing area, and storage. The space must have:

• Good ventilation or air conditioning (operators working in heat produce less and leave sooner)
• Stable power supply with backup (a power cut that stops 40 machines for 2 hours costs you 80 man-hours)
• Adequate lighting (minimum 500 lux at machine level — poor lighting causes eye strain and defects)
• Clean restrooms (this is not optional — it is a labor compliance requirement and basic dignity)
• Ground floor access for cutting (fabric rolls are heavy, do not carry them up stairs)

3. Machines

The exact machine mix depends on what you are making. For a general CMT factory doing basic knit garments (t-shirts, polo shirts, joggers), here is a practical machine list for 40 operators:

Total new: approximately NPR 9-10 lakh. Total used: approximately NPR 4-5 lakh.

My advice: buy used machines for everything except buttonhole and button attach. These specialty machines need precision, and worn-out ones create quality problems. For overlock and flatlock, a good used Japanese machine (Juki, Pegasus, Yamato) will outperform a new Chinese machine at half the price. Check the needle bar play, feed dog condition, and looper timing before buying.

4. People

You need the following key people from day one:

• 1 production supervisor per 20-25 operators — This person runs your line. They allocate work, manage operators, handle quality issues, and keep the line balanced. A good supervisor is worth more than any machine. Hire someone with at least 3-5 years of experience in a garment factory.
• 1 cutting master — Cutting is where quality starts. A bad cut means every garment on that layer is defective. Your cutting master needs to understand marker layout, fabric spreading, and layer cutting. This is not a job for a trainee.
• 1 quality checker — Someone who checks garments at inline and end-of-line stages. They need to know construction standards, measurement tolerance, and common defects. Many small factories skip this role and have the supervisor do quality checking. That is a mistake — the supervisor is busy managing the line and will always prioritize output over quality.
• Operators — Start with experienced operators who can work independently. Do not start a factory with 40 trainees. Mix experienced and trainee operators at a ratio of about 70:30. Experienced operators set the pace and help train the others.

5. Legal registration

In Nepal, you need: a company registration certificate, PAN/VAT registration, factory registration from the Department of Industry, and labor compliance documentation. The specifics vary by country, but every country requires some form of business license and factory inspection before you can legally operate. Do this first — getting shut down two months in because you skipped registration is expensive and embarrassing.

Finding Your First Order

This is the part that nobody talks about honestly. Finding orders is harder than setting up the factory. You can buy machines in a week. Finding a buyer who trusts you with their order takes months.

Here is what works, based on my experience:

Start local

Do not chase export orders when you start. You do not have the compliance certifications, the production track record, or the relationships. Start with local brands, local traders, and local retailers who need small quantities made. In Nepal, there are dozens of small clothing brands that outsource production. They order 500-2,000 pieces at a time. The margin is decent because there is no export overhead.

Build a sample room first

Before you set up the full factory, set up a sample room with 3-4 machines and your best operators. Make samples for potential buyers. A buyer will not give you a production order based on a conversation. They will give you a production order based on a sample that meets their quality standard. Your sample room is your sales tool.

Word of mouth is everything

In the garment industry, especially in South Asia, relationships matter more than marketing. One good delivery leads to a referral. One late delivery kills three referrals. Your reputation is built one order at a time, and it is your most valuable asset. Deliver on time, at the right quality, and do not negotiate on price after you have agreed.

Visit garment clusters

Every country has garment manufacturing clusters — areas where factories, fabric traders, and buyers congregate. In Nepal, it is the Kathmandu valley and certain areas in the Terai. In Bangladesh, it is Dhaka and Chittagong. In India, it is Tirupur, Ludhiana, and Noida. Visit these clusters, talk to people, learn who needs manufacturing capacity. Buyers in these clusters are always looking for reliable CMT partners.

The First 90 Days

Here is what the first three months actually look like. Not the plan — the reality.

Week 1-2: Setup and chaos

You install machines, set up the cutting table, run electrical wiring, and realize the ventilation is worse than you thought. Half the machines you bought need servicing. The mechanic you hired does not show up on day two. You call another one. The cutting table is 30 centimeters too wide for the space. You adjust.

You hire operators. Some of them do not come back after the first day because the factory is too far, the pay is not what they expected, or they found another job. You hire replacements. You start to understand why experienced factory owners always hire 10% more than they need.

Week 3-4: Pilot production

Your first real production run begins. You have confirmed an order for 2,000 polo shirts. The buyer has sent fabric. Cutting starts. Your cutting master lays the fabric, cuts the first batch, and you discover the marker layout wastes 8% more fabric than the buyer expected. You redo the marker. Cutting starts again.

Sewing begins. Your experienced operators hit their stride within a day. Your trainees are slow and need constant supervision. The line produces 200 pieces on day one instead of the 400 you planned. Your supervisor tells you the operation breakdown needs adjusting because two stations are bottlenecked. You rebalance.

Week 5-6: First quality rejection

The buyer's QC person visits and rejects 15% of the first batch. The collar attachment is uneven. The side seam measurement is off by 1 centimeter. The thread color on the topstitch does not match the approved sample. Your stomach drops.

You fix the issues. Your quality checker starts catching these problems before they reach the buyer. You learn that quality is not something you add at the end — it is built into every operation. You add inline checks at collar and side seam stations.

Week 7-8: First delivery

You deliver the first order. It is three days late, but the quality is acceptable after the rework. The buyer is not thrilled about the delay but says they will give you another order if you can deliver on time. You promise you will. You mean it.

Week 9-12: Finding rhythm

The second order goes smoother. You have learned which operators work best on which machines. The supervisor has figured out the optimal line balance for this type of garment. Cutting is faster because the cutting master knows the fabric now. You deliver on time. The buyer places a third order, larger than the first two.

Somewhere in this period, you also have your first payment dispute. An operator claims they completed 500 pieces this month. Your records show 420. The argument takes two hours. You realize you need a better system for tracking who did what. WhatsApp messages and notebook tallies are not going to cut it.

When to Invest in Software

I used WhatsApp groups and Excel spreadsheets for the first three months. It worked. Barely. I had a WhatsApp group for each line, and the supervisor would send production counts at the end of each hour. I tracked orders in Excel. I calculated operator payments by hand.

It worked until we had three simultaneous orders running on three lines. Then I lost track of which bundles were where. An operator's payment was miscalculated because the supervisor wrote "450" instead of "350" in the WhatsApp message, and I did not catch it until the operator confronted me. A supervisor quit over a payment error that was entirely my fault — I had double-counted a batch in the Excel sheet.

That is when I built Scan ERP. But I do not think you need software on day one. Here is my honest timeline:

• 0-3 months: Excel and WhatsApp are fine. You have one or two orders and a small team. You can manage manually.
• 3-6 months: You start feeling the pain. Payment disputes, lost bundles, unclear WIP counts. Start looking at production tracking software.
• 6+ months: If you are running more than two lines or more than three simultaneous orders, you need a system. Manual tracking at this scale creates errors that cost more than the software.

The goal of production software is not to add complexity. It is to remove the errors and guesswork that multiply as you grow. When an operator scans a QR code instead of relying on a supervisor's handwritten tally, payment disputes disappear. When you can see WIP at every station in real time instead of waiting for end-of-day counts, bottlenecks get fixed in minutes instead of hours.

What I Would Do Differently

If I could start over, here is what I would change:

Start smaller

I started with 50 operators. I should have started with 20. Twenty operators on one line is manageable. You learn the operations, you learn management, and you build quality consistency — all without the cash burn of running 50 operators through the learning curve. Scale to 40-50 once your first line is running smoothly and you have consistent orders.

Get the cutting room right first

Cutting is the foundation of everything that follows. A bad cut cascades into quality problems at every sewing operation. I underestimated cutting and overfocused on sewing. Invest in a good cutting master and proper cutting equipment from day one. A well-cut bundle makes every downstream operation faster and cleaner.

Do not hire a merchandiser too early

A merchandiser is the person who interfaces with buyers, manages orders, and coordinates production planning. They are essential in a factory with 5+ regular buyers. But when you have 1-2 buyers, you can do this yourself. Merchandisers cost NPR 30,000-50,000 per month. Save that money until you have at least three regular buyers and the coordination becomes too much for you to handle alone.

Build relationships before building capacity

I bought machines before I had enough orders to keep them running. For the first two months, 30% of my machines were idle because I did not have enough work. That is NPR 1-2 lakh per month in rent and overheads wasted on idle capacity. The correct order is: find buyers, confirm orders, then add capacity to match. Not the other way around.

Pay operators on time, every time

Late payment is the fastest way to lose good operators. In Nepal's garment industry, skilled operators are in short supply. If you delay payment by even a week, your best operators will leave for the factory down the road that pays on time. I learned this the hard way. Now our payment processing happens within 48 hours of month-end, every month, no exceptions.

Starting a garment factory is not glamorous. It is not a tech startup story. It is daily problem-solving on a factory floor with real people, real machines, and real deadlines. But if you get the fundamentals right — confirmed orders, good people, proper layout, and relentless attention to quality — it works. The margins are thin, but they are consistent once you build your reputation. And there is something deeply satisfying about watching raw fabric turn into finished garments on a floor that you built.

Santosh Rijal is the founder of Scan ERP, a garment manufacturing ERP system designed for factory floor operations. He is also a medical doctor who decided factories were more interesting than clinics. He works with garment manufacturers across Nepal.