$15+ is the average handling cost per returned garment, but its effect varies by apparel category. Before adopting fit recommendations built from two phone photos, a brand needs a credible estimate of the return savings. Return reduction is measured later, pilot data needs time, and the inputs differ enough that one benchmark cannot represent every brand.
This is the bottom-up framework we use with brands assessing phone-photo fit recommendations. It relies on figures operators already track or can retrieve from their analytics stack within a day, rather than a top-down benchmark. The examples are illustrative. Your return and cost data should drive the model.
Three inputs drive the model
The core ROI calculation uses three inputs. First, measure the current apparel return rate as a share of items shipped. Next, identify the share of returns caused by sizing or fit, excluding defects, transit damage, and wrong items. Finally, calculate cost per return from outbound shipping, covered return shipping, receiving labor, inspection, and any reprocessing needed before resale.
Resale discount rate is a fourth input that can materially change the result. Compare returned items sold at full price with those requiring markdown or liquidation. Strong resale rates lower the economic cost of a return. Fit returns that arrive worn, altered, or outside the resale window cost more, making the case for phone-photo fit recommendations stronger.
Build the base case
Consider a hypothetical brand shipping 10,000 items monthly at an average order value of $85. With a 22% return rate, 2,200 items come back. If 60% are attributed to sizing or fit, that produces 1,320 size-related returns monthly. The all-in return cost is $18: outbound label $8, return shipping $6, and receiving and inspection $4. Of fit returns, 40% need a $15 average markdown. Monthly cost is 1,320 x $18 = $23,760 in handling, plus 1,320 x 40% x $15 = $7,920 in markdowns, totaling $31,680.
If phone-photo fit recommendations cut size-related returns by 40%, the Q4 2025 pilot cohort figure measured over 90 days, this brand saves 40% x $31,680 = $12,672 monthly. At $199/month for the Brand plan, net monthly savings are $12,473. ROI is positive in month one if the pilot result applies to this product mix.
Where the model shifts
The example is a middle case. Garment construction can move the economics substantially. Fitted wovens, structured outerwear, and tailored trousers usually bring more fit returns and higher inspection costs, while returned structure shows wear and resells less easily. Fit recommendations are therefore more valuable in these categories. Stretch knits and casual basics generally produce fewer fit returns, so per-SKU ROI is weaker, though total volume can still matter.
Adoption among shoppers is the other major variable. A fit recommendation shown at size selection and completed quickly can reach a meaningful share of shoppers. One that sends shoppers away from the product page, requires an app download, or uses a long quiz will see limited use. Our embed reduces those steps, but adoption still belongs in the model. If 30% use it, the modeled reduction applies to 30% of shoppers, not the full volume.
Conversion upside is harder to model
Fewer returns are the primary economic case. Conversion is a secondary benefit that is harder to forecast but appears in pilot data. A shopper who gets a confident fit recommendation while choosing a size may purchase instead of abandoning because the fit is unclear.
We see a conversion effect in pilot data, but do not publish one universal lift because it depends on PDP quality, abandonment at size selection, and garment category. A brand with high size-step abandonment and a fitted, size-complex assortment may see more lift than one whose shoppers already know their size. We measure each partner's result in context instead of applying a generic benchmark.
What the model leaves out
The direct model covers logistics and markdown costs, not downstream customer lifetime value. A poor fit can make a shopper less likely to buy again, while fit-related returns can raise acquisition cost relative to repeat revenue. Improving the first order through a 3D body model and garment fit recommendation may improve LTV, but isolating one intervention's effect is methodologically difficult.
The model also omits the product insight in fit data. Comparing shopper body measurements from phone-photo body models with garment specifications creates measurement distributions for the actual shopper base. Brands can use them for grading, size-range priorities, and extended-size photography briefs. We view this as a byproduct, not a primary ROI input, because its future value is difficult to quantify.
Run the model for your brand
Collect monthly units shipped, current overall return rate, the share attributed to fit or sizing, all-in cost per return, and the average markdown on fit returns that do not sell at full price. Operations or finance should track these figures. If fit attribution is not separated, customer service likely holds it in return-reason codes, even without a clean aggregate report.
Run three cases: 25% return reduction for a conservative view, 40% for the median, and 55% for the optimistic case. A meaningful positive result in the conservative case makes a pilot easier to justify. If the optimistic case is still marginal for your volume and mix, fit recommendations may not be the highest-impact option, which is useful to know before committing.
We can run this model with your actual inputs. In practice, brand-specific results are often stronger than generic benchmarks suggest. To review your numbers, start on the contact page.
