Sourcing from Pakistan is a specification problem before it is a supplier problem
Every buyer arrives at this point believing the hard part is finding a factory. It is not. Pakistan has thousands of garment manufacturers, and a buyer with a clear specification can usually get three credible quotations inside three weeks. The hard part is that most buyers arrive with a sentence — “400gsm hoodies, good quality, about a thousand” — and a sentence cannot be quoted. It can only be interpreted.
Interpretation is where money quietly leaves your control. Two factories reading the same sentence will build two different garments, return two different prices, and both will be telling the truth about the thing they thought you asked for. The comparison you then make is not between suppliers. It is between two different products.
So the sequence below starts before the factory, and the whole of it is about making your own requirement unambiguous. That is the work that decides the outcome.
Start with the category, because it decides the city
Pakistan’s apparel export base is concentrated, and the concentration is geographic as much as it is product-based. Knitwear, readymade garments and bedwear are the largest single export categories, and they are not made everywhere.
| Product | Strongest base | Why |
|---|---|---|
| Knitwear, jersey, tees, hoodies | Karachi, Lahore, Faisalabad | Vertically integrated knit-and-dye capacity; the largest export volume sits here |
| Denim and woven bottoms | Karachi, Lahore | Specialist laundries and finishing, which is where denim is actually won or lost |
| Bedwear and home textiles | Karachi, Faisalabad | Long-established export category with its own machinery base |
| Woven shirts, formalwear | Lahore, Faisalabad | Large woven capacity, though shirts are a more competitive global segment |
| Sportswear, gloves, technical | Sialkot | A distinct cluster with its own history and skills base |
| Fabric and yarn supply | Faisalabad | The spinning and processing centre of the country |
Two things follow from this table. First, the city is a proxy for capability, and a factory claiming strength outside its city’s specialisation is a risk to test rather than a bargain to take. Second, and more importantly: choosing the category chooses which set of constraints you are going to be managed by — a knitwear factory’s problems are dyeing-lot and shrinkage, a denim factory’s are washing and hand-feel, a woven factory’s are loom availability and grey fabric. Buyers who understand which family of problems they bought get far fewer surprises.
Then decide the three things that set the floor price
Price on a garment is not primarily a negotiation. It is mostly the residue of three decisions taken before any negotiation begins.
Fabric. In the ILO’s own analysis, fabric is, as a rule, the most costly item in the overall FOB price of a garment [5]. The Better Work country study puts materials plus direct labour together at 55–65% of FOB across the countries it examined [4]. That means the factory’s own margin, and everything you might argue about, sits in the minority of the number. If you want the price to move materially, the specification is the lever, not the conversation.
Construction and finishing. The difference between a garment that survives twenty washes and one that does not is usually a finishing step, a stitch density, or a tolerance — each of which costs money that does not appear in a headline quote. Specifying “no visible shrinkage beyond 5% after three washes” is a sentence that changes who is responsible when it fails.
Order structure. Colour count, size curve, and how many separate SKUs you split the quantity across drive cutting-room efficiency and dyeing economics. A single colour at 1,000 pieces and four colours at 250 pieces each are not the same order to a factory, whatever the total reads.
Know the cost structure, then decide who carries each risk
Pakistan’s industrial cost base today is a mixture of genuine competitiveness and genuine friction. Labour remains low: the notified minimum wage for unskilled workers in Punjab and Sindh is PKR 40,000 per month, effective from 1 July 2025, on a 26-day month [7]. Energy is the opposite story — the industrial tariff cited in NEPRA’s January 2026 decision is around US¢12.90 per kWh [6], which is a direct competitive disadvantage against regional competitors and a reason power-intensive processing steps are priced sensitively.
That combination produces a specific and predictable failure mode: buyers negotiate hard on the labour element, which is small, and leave the fabric, energy, finishing and risk allocation elements unexamined, which are large. The table below is the set of decisions that matter more than the price argument.
| Decision | Option A | Option B | Who carries the cost when it goes wrong |
|---|---|---|---|
| Who owns the specification | You supply the tech pack | Factory supplies the design | A: you. B: the factory, and so does your competitor |
| Fabric sourcing | Buyer-nominated mill | Factory’s own mill | A: you, including the delay. B: the factory, with less visibility |
| Tolerance | Written and testable | “Industry standard” | A: defined. B: whoever is louder when it fails |
| Inspection | Third-party, pre-shipment | Factory self-inspection | A: you pay, you see first. B: you learn at your warehouse |
| Payment trigger | Documented milestones | Calendar dates | A: progress. B: elapsed time |
| Freight and duty | Agreed incoterm, buyer-arranged | Factory-arranged to door | Whoever accepted the incoterm in writing |
There is no universally correct column. There is only the question of whether you have chosen deliberately, and whether the choice is written down.
The sequence, in the order that avoids rework
- Write the specification. Product, fabric composition and weight with a tolerance, construction, finish, colour references, embroidery or print with placement and size, packaging, packing ratio, and the tests you will actually apply.
- Decide the commercial model. OEM or ODM changes who owns the pattern and whether you can move production later. This is covered in full at /sourcing/oem-vs-odm/.
- Choose the city from the product, using the table above.
- Approach three to five factories with one document, so quotations are comparable. Ask each one what they would change about your specification and why — the answers are more informative than the prices.
- Confirm the sample against a written standard, not against how it looks. Record measurable attributes so the approval means something at the end of a production run.
- Agree inspection, tolerances and remedy before the order, not after a problem. The moment to negotiate what happens when 4% of a run fails is while the order is still worth something to the factory.
- Tie payments to milestones you can verify. Production photographs, inspection reports, loading documents. Not elapsed time.
- Book freight and confirm duties yourself until you know exactly what your factory is quoting on your behalf.
Where buyers lose money, in the order of how often it happens
Under-specified briefs, which produce incomparable quotations and, later, a dispute neither side can resolve from the documents.
Sample approval by feel, which makes the sample a conversation rather than a standard, and leaves the golden sample unusable as evidence.
Unallocated risk. The single most expensive sentence in sourcing is “we agreed that verbally”. Everything above is written down for this reason.
Duty exposure treated as fixed. The US position on Pakistani goods has changed repeatedly since 2025, having moved from a 29% reciprocal tariff in April 2025 onwards [9]. A landed-cost model built on a rate from last quarter is a model built on nothing.
What you are buying when you buy from Pakistan
Pakistan’s advantage is not that it is the cheapest place to make clothing. Bangladesh is larger in garments and China is larger still, and both beat Pakistan on particular orders. Pakistan’s advantage is depth in cotton-based vertical production — spinning, weaving, processing and stitching in the same country, often the same company — plus the ability to run mid-sized orders without a buyer having to be a global brand to be worth someone’s attention.
That is a real advantage and a narrow one. It survives when you specify well, and it disappears when you specify badly, because a specification failure is the one cost no country can make cheap.
Worked examples
Worked example: the same hoodie, two specifications, 41% apart
A 400gsm brushed-back fleece hoodie at 1,000 pieces will quote very differently depending on decisions the buyer thinks are cosmetic. Using Pakistani knitwear pricing structure: a 320gsm single-jersey-based hoodie in 100% cotton with a basic embroidery hit and polybag packing sits at one price. Moving to 400gsm brushed-back fleece with a heavier yarn, a three-colour puff print, a woven label and a fitted carton raises fabric consumption by roughly a quarter, adds a finishing step and changes the packing. On the same order quantity, the second specification is approximately 41% more expensive per piece — and neither buyer would describe the difference as "a different product". The number to take from this is not 41%. It is that the specification, not the factory, moved the price.
Worked example: why the third quotation is not the cheapest factory
Send the same brief to five Pakistani factories and the returns usually cluster in two groups rather than five prices. The lower group is typically quoting the garment you described in the words you used; the higher group is quoting the garment you meant, and has priced the tolerance, the testing and the inspection that the lower group left out. Buying on the lowest number therefore does not buy the same thing more cheaply — it buys a different promise. The correction is not to pick the higher price; it is to send a single specification to every factory so the comparison is real, and then ask the low quoter what happens at the tolerance you actually need.