Accurate size recommendations need more than body measurements. They need fashion knowledge: how sizes differ between brands, how a cut is meant to fit, and how real bodies differ from a formula. Selectika builds size guidance on that knowledge, so shoppers choose the right size the first time.
What a sewing workshop taught me about size
I grew up among the sewing machines in my grandmother's workshops. She was a fashion designer, and in her world, nobody trusted the number on a label. They trusted the tape measure, the fitting, and their own eyes.
Almost none of that knowledge was written down. Where a waist should sit. How much ease a jacket needs so you can lift your arms. Why two women with the same measurements can need different sizes in the same dress.
Years later, working on pattern design and couture draping at Sonia Rykiel in Paris, I saw the same thing from the inside. A size isn't a fact about a body. It's a set of decisions a pattern maker made about one imagined body, and every brand makes those decisions differently.
Why is choosing a size online so hard?
Choosing a size online is hard because sizes are not standardized. A medium in one brand can fit like a small or a large in another, and even within one brand, a slim cut and a relaxed cut in the same size fit very differently. Fabric matters too: a stretch knit and a stiff linen behave nothing alike.
In a shop, you solve this in the fitting room. Online, the shopper guesses. Many order two sizes and send one back, and wrong sizes are among the most common reasons clothes are returned.
Why do many size tools get it wrong?
Many size tools rely on simple body formulas such as BMI, built from height and weight alone. BMI says nothing about where a body carries its weight, how long the legs are, or how broad the shoulders are, and it knows nothing about the garment. A formula like that can be precise and still wrong.
How does Selectika recommend sizes?
Selectika recommends sizes the way an experienced fitter thinks, not the way a formula does. The approach rests on four principles:
- Real measurements, not BMI. The engine works in centimeters, using body measurements instead of a single height-and-weight score.
- Start from what the shopper already knows. A shopper can start from the size they wear in a brand they know, and the engine translates it to the brand they're viewing, because brands size differently.
- Body shape, from fashion experts. The rules for how different body shapes fit into garments were built with people who have fitted real clients, not only from data.
- Honest about uncertainty. When the answer sits between two sizes, the engine accounts for that instead of pretending to be certain. A good fitter would say the same.
The engine also reads each garment's attributes, such as fit and fabric, so a slim-cut shirt and an oversized one get different advice.

Shoes: foot length and brand conversion
For shoes, Selectika recommends sizes from foot length and converts between brands' size systems, so a shopper who knows their size in one brand can find it in another.
When the right size is sold out
Getting the size right is only half the job. When the shopper's size is sold out, Selectika's Find In Your Size shows similar items that are in stock in that size, right on the product page, so the visit doesn't end there.

What does this mean for retailers?
Accurate size recommendations mean fewer returns, fewer double orders, and shoppers who buy with confidence.
Size guidance also works best alongside the rest of Selectika's fashion AI: complete outfits, color-aware recommendations and style personas all run on the same catalog knowledge. Read: The Salesperson Your Online Store Is Missing
Turning workshop knowledge into data
My grandmother never wrote down how a sleeve should fall. She didn't need to; she could see it. Online stores can't see, so that knowledge has to be translated into something a system can use, carefully, by people who understand it.
That's the work we do at Selectika: taking what designers, fitters and stylists simply know, and building it into the systems behind online stores.
Get size recommendations that understand fashion
Fashion retailers: See Selectika's size guidance and Find In Your Size running on your own catalog. Book a demo →
Investors: We're building the fashion intelligence layer for online retail. We'd be glad to talk. Contact →
About the author
Nuphar Gabrieli is a fashion designer who has worked on fashion AI at Selectika since 2020. She holds a B.A. in fashion design and an M.A. in knitwear and menswear from Atelier Chardon Savard in Paris, worked on pattern design and couture draping at Sonia Rykiel, designed and bought menswear for Renuar, and co-founded her own label.
Frequently asked questions
Why are clothing sizes different between brands?
Clothing sizes aren't standardized. Each brand builds its patterns around its own idea of a typical body, so a medium in one brand can fit like a small or a large in another. Cut and fabric change the fit further, even within one brand.
Why isn't BMI enough for size recommendations?
BMI uses only height and weight. It says nothing about body shape, proportions or how a garment is cut, so two shoppers with the same BMI can need different sizes.
How does Selectika recommend clothing sizes?
Selectika's size engine uses body measurements in centimeters instead of BMI, lets shoppers start from a size they know in another brand, applies body-shape knowledge built with fashion experts, and accounts for uncertainty when a shopper is between sizes.
How does Selectika recommend shoe sizes?
Selectika recommends shoe sizes from foot length and converts between brands' size systems, so shoppers can find their size in a brand they haven't worn before.
What happens when a shopper's size is sold out?
With Selectika's Find In Your Size, the product page shows similar items that are in stock in the shopper's size, so the visit continues instead of ending.
How do accurate size recommendations help fashion retailers?
Accurate size recommendations reduce returns and double orders, because shoppers choose the right size the first time and buy with more confidence.

