As of 1 November 2026, Boody has published the results of a sizing experiment that most apparel brands would rather not run. The company found that 41 per cent of its customers cannot or will not use a soft measuring tape, so it stopped asking them to. Working with Magic Fit since March 2026, it built a guided sizing experience and then tested it properly: across 230,000 sessions from August 28, shoppers in Australia and New Zealand were split 50/50 between the existing size chart and the guided version. About one in eight new customers in the guided group would not have bought at all, and first-time buyers reordered 19 per cent more often over the following 90 days.
41 per cent of shoppers will not use a tape measure
The number that matters is not the conversion lift, it is the reason behind it. Boody sells wire-free bras that run from XS to 4XL across three cup ranges, which is a size language most shoppers have never learned. A standard chart asks the customer to translate her usual band-and-cup size into that language using a tape measure she may not own. The guided experience removes the translation step, and the result shows how much revenue was sitting in that gap.
Extrapolated over a year using the company’s own cohort history, the guided sizing is worth an extra $700,000 in annual revenue in Australia and New Zealand alone. That is a large number for a change that does not touch the product. It is also a reminder that in apparel, the product page is part of the product: if the page cannot help a shopper choose a size, the range effectively does not exist for her.
What the split test showed
Bounce, not delay
The distinction the company draws is important. Customers in the control group were not delaying a purchase or switching to another product, they were leaving. A sizing failure at the top of the funnel is usually read as a conversion problem, which sends teams towards discounting or retargeting. The test suggests it is a fit-confidence problem, and discounting does not fix confidence.
Reorder lift is the durable effect
The 19 per cent reorder lift among first-time buyers is the more valuable finding, because it persists after the first order. A shopper who gets the size right the first time learns the brand’s size language and stops needing help. That turns a one-off assist into a repeat customer, which is why sizing work belongs with product development rather than with marketing campaigns.
There is a product concept question underneath this. The size language, the cup structure and the labelling all express one consumer need, which is to buy underwear without measuring anything. If the product concept assumes a shopper who knows her band and cup size, the range is designed for a customer the brand does not have. That is why the fix landed on the product page rather than on packaging alone. A brand that runs three cup ranges across a wide size span carries more SKU combinations than a brand with a simpler size system. If guided sizing lifts conversion without changing return rates, the cost of that complexity falls, and the case for keeping the full size range gets stronger. If returns rise at the same time, the extra combinations are being sold to the wrong shoppers.
Three columns for sizing tests
Most sizing projects fail because they are measured with conversion alone. The table below is the minimum set of columns for judging whether a guided sizing change is real, and it can be built from a standard analytics stack. Read together, the three columns answer a single question: did the tool help shoppers choose, or did it help the warehouse clear stock that would otherwise have been returned anyway.
| Column | Definition | Warning sign |
|---|---|---|
| Bounce at size step | Exit rate on the size selection page | Unchanged after the change |
| Return rate by size | Returns divided by orders for each size | Rising for the sizes the tool recommends |
| Reorder rate | Repeat orders within 90 days | No lift among first-time buyers |
The second column is the one that catches false wins. A guided tool can lift conversion simply by pushing shoppers into the sizes that are in stock, which shows up later as a return spike and a margin hit. Reading the return rate by size next to the reorder rate tells you whether the tool is helping shoppers choose or helping the warehouse clear.
FAQ
Why did a size chart problem cost so much revenue?
Because the chart asked shoppers to translate their usual size into a brand-specific language using a tool many do not own. The guided version removed the translation step, and roughly one in eight new customers would not have bought without it.
Is the conversion lift the most important result?
No. The 19 per cent reorder lift among first-time buyers is more durable, because it persists after the first purchase. Conversion can be bought with discounts, reorder lift cannot.
Could guided sizing increase returns?
It can, if the tool nudges shoppers towards whatever is in stock. That is why return rate by size has to be read alongside reorder rate; a conversion gain with a return spike is not a win.
Does this change how size ranges are planned?
It can. If guided sizing lifts conversion without lifting returns, the cost of carrying a wide size range falls, which strengthens the case for keeping the full range rather than trimming it.
