CollectedAI
Brands
Factory costs

What robots do for a clothing brand

Profit per garmentWorkersRobots
Brand
Automation progress

Where the price goes

Every input is sourced or marked as our estimate.

Brand

From the brand's filings, or our estimates where it doesn't report them. They reset when you switch brands.

Set by the stage above. Move any dial to try your own.

Useful output while running, including waits for work, compared with a trained worker sewing at a steady pace.

A technician is on duty whenever robots run, at least one shift a day. Assumes factories big enough to keep each technician busy.

Office, building, and support costs, compared with a traditional factory.

Measured in days of the robot's own output. A sewing line loses about one day. Matters most for small orders.

Added US tariff on Asia

Fabric for US and Mexican factories

Costs that depend on where the work is done.

Short-run premium from human factories (%)

Worker wages, with benefits ($/hr)

Custom fit

Robot factory

At a glance

Takeaways

    Each takeaway is computed from the model and tested on random mixes of every input's plausible range.

    What matters

        • Made-to-measure brands. The custom fit dial covers them.
        • Uniforms. Covered by the garment model, where US law leaves US factories as the only rivals.
        • Switching costs. Moving production takes time and money: new suppliers, factory build-out, and capital for robots. Results show the brand once it has switched.
        • Fabric supply. Mexican factories need North American fabric to ship to the US duty-free. We assume Mexico's mills grow to supply it, and that fabric is stocked ahead so it never slows nearby factories.
        Check it yourself

        Each bar moves one assumption across its plausible range and shows the margin points the brand gains. Longer bars matter more.

        How the model works

        The brand keeps its prices. For each way of making its core item, the model works out the profit on each one sold in the US and picks the best.

          Assumptions

          Every input is sourced or marked as our estimate.

          Dialyou set it Sourcedfrom data Estimateour judgment

          Sources, by question

          Open any question to see its sources and the figure each one gives.

          The brands

          What does a brand like Old Navy earn, sell, and make where? Gap Inc. earns a 7.3% operating margin on $15.4 billion of sales (FY2025). Old Navy sells $8.7 billion; its $7.6 billion of US sales is our estimate, using Gap's 88% US share. Nearly all of it is made in Asia, mainly Vietnam and Indonesia. 3 sources
          What does FIGS earn, sell, and make where? A 6.0% operating margin on $631 million of revenue (FY2025), $528 million of it in the US. It makes its scrubs in Vietnam and Jordan, from fabric mostly made in China. 4 sources
          What does a brand like Shein earn, sell, and make where? A 4.1% operating margin on $41.9 billion of revenue (2025). Its US sales of about $9 billion are our estimate from a 22.5% to 24% US share. It makes in China through about 7,500 contract makers, and its knit tops now pay 52% US duty. Its figures come from reports of its prospectus. 5 sources
          What does a brand like Levi's earn, sell, and make where? A 10.8% operating margin on $6.3 billion of revenue (FY2025), with about $2.7 billion, or 42%, in the US. It uses contract makers in about 32 countries, none supplying more than 30%. 4 sources
          What does an emerging brand earn, sell, and make where? Our estimate for a small US brand making short runs in Asia: a 3% operating margin on $20 million of US sales, with 90 days of stock. Small public brands run from a 4.5% loss to a 6.1% profit, and benchmarks for brands under $10 million cluster at 0% to 5%. 5 sources
          What does each brand receive for its core item? List prices: a $15 Old Navy tee, a $94 FIGS set, a $10 Shein top, $69.50 Levi's 501s, and a $35 emerging-brand tee. No brand discloses what it actually receives, so the shares of list (60%, 90%, 90%, 62%, and 85%) are our estimates, as are the Shein and emerging-brand prices. 5 sources
          How much of a brand's revenue goes to making its clothes? Old Navy 47.1%, FIGS 33.5%, and Levi's 38.3%, from filings, including duty and inbound freight. Shein's 37% of its product revenue is our estimate and covers the factory price only, since Shein appears to book duty and freight elsewhere. The emerging brand's 35% is our estimate from small public brands. 5 sources

          Speed and stock

          What does guessing demand wrong cost? Making months ahead in Asia, unplanned markdowns, write-offs, and lost sales cost 3% to 4.5% of revenue for Old Navy, FIGS, and Levi's, 4% for Shein, and 7% for an emerging brand. These blend normal and bad years from filings; the jeans, Shein, and emerging-brand figures are mostly our estimate. 5 sources
          How much does making clothes nearby cut that? Nearby workers cut it by about 40% to 50%, to 1.8% to 2.5% of revenue for the large brands, and robots by about 55% to 65%. Shein, already fast from Asia, gains half a point to a point. The ratios are our estimate from a forecasting model; robots win mostly on cost, not speed. 5 sources
          How much stock do brands hold, and what does holding it cost? From 36 days (Shein) to 210 (FIGS), from filings; the emerging brand's 90 is our estimate. Making nearby removes about 45 days of goods in transit and safety stock, or 10 for Shein, which flies its goods. Storage, insurance, and shrink cost about 3% of stock value a year; capital, at 7%, is shown as cash freed, not as profit. 5 sources
          What do human factories charge for short runs? At about 300 units a style, about 50% more in Asia, 35% in Mexico, and 40% in the US. Less the learning cost the model already charges, we add 45%, 30%, and 30%. Shein's suppliers charge almost nothing extra, because reorders repay them, and robots pay only on the work people still do. 5 sources
          Do reorders from nearby factories help? Yes. Making 30% of stock nearby as reorders, once early sales show what sells, gets a brand 60% of the benefit of making everything nearby. Both are our estimates: 30% sits between H&M and Zara, and 60% is below our simulation's 75% to 90% to allow for sizes, minimums, and fabric. 4 sources

          Robots at the brand level

          How busy would a robot factory be? 80% of the year when shared by many brands and 50% when one brand owns it. These are our estimates; apparel demand peaks in December for everyone, so sharing alone tops out near 67% to 77%, and the rest comes from making basics ahead in slow months. 5 sources
          What would custom fit cost and earn? Our estimate, offered only from robot factories: 15% of buyers pay 25% more, each custom garment costs 22% more to make, and 5% are remade. It is off by default; jeans are the one category with measured demand. 5 sources
          How much of a brand's savings reaches shoppers? None by default, so results show what a brand gains while only it has robots. You can raise it: history suggests 60% to 90% once rivals have the same robots (our estimate). 5 sources
          Who sets prices in each category? Scaled specialists, not typical brands: Gildan and private-label multipacks for tees, Shein and Zara for tops, and Walmart and FIGS for scrubs. We found no comparable check for jeans. 5 sources

          Garments

          Where do the factory costs come from? From the first model, What Robots Do to the Cost of Clothing, and its sources. We scale each brand's item so that making it where the brand makes it today matches the brand's reported product cost. 3 sources
          What does a pair of jeans take to make? 21 minutes of work, 67% of it sewing; washing and hand finishing add the rest. Materials, wash included, cost $6.50 in Asia and $9.00 in North America, the latter our estimate. Duty is 16.6%, a pair weighs 0.70 kg, and 8% goes unsold, also our estimate. 5 sources