For many online fashion retailers, returns are the Scrooge of the holiday season.

Not only are they costly in terms of reverse logistics, but often the items are already marked down by the time they’re ready to go out again, if they can even be resold at all. And, as we see returns skyrocket in late December and early January, some could surely be avoided if only the industry could solve the issue of fit.

While the overall return rate for e-commerce purchases is around 30 percent (compared with 10 percent in stores), it can be even higher for retailers that sell clothing and accessories. Zalando, Europe’s largest online fashion retailer, sees 50 percent of all merchandise sold eventually sent back, eating into already-thin margins. Reasoning that customers would be less inclined to make returns if everything fit correctly on the first try, the company has enlisted a team of models to try on shoes and dresses as they go live on the site and report back about sizing and other issues.

According to a recent Reuters report, Zalando’s models try on up to 120 pairs of shoes per day, with each pair tested by three models to account for subjective opinions around fit.

Asos launched its Fit Assistant tool in November in collaboration with the Berlin-based software firm Fit Analytics, whose machine-learning platform also powers sizing recommendations for brands like Mango, Puma and Tommy Hilfiger.

“We want to do all we can to make sure our customers are getting the right size, first time,” said Andy Berks, digital product director at Asos. With the new tool, he said, “we can now offer personalized size recommendations across Asos collections, exclusive labels and fashion favorites, helping make our customers’ lives easier and their shopping experience even better.”

Brands including Aldo, DSW, Ugg and Cole Haan work with data-driven personalization platform True Fit to help increase conversion and reduce returns; in the five days between Thanksgiving and Cyber Monday, True Fit said it saw 10.5 million daily users and a total 2.7 million checkouts, an 85 percent year-over-year increase. Customers that used the tool also checked out at twice the rate of those that didn’t, the company said.