Choosing a size recommendation tool is a decision that affects the product page, your catalogue data and the work of the e-commerce team. The demos all look alike: a shopper enters a few details and a size appears. The differences lie in what the demo does not show.
These are the questions to ask every vendor before choosing, us included. Ask for written answers so you can compare them.
How the calculation works
- What data does it ask the shopper for? More fields mean more friction, fewer fields may mean less accuracy. Ask which data is required and what happens if the shopper does not know it.
- How does it reach a size? A mathematical algorithm, a model trained on purchase data, or a mix of both. There is no single right answer, but you should know what you are buying and be able to explain it.
- What does it compare the shopper with? Each garment's measurements, a generic chart per category, or return history. A tool that works on garment measurement charts takes into account fit differences between one style and another.
What you need to provide
- What product data is needed? Measurement charts, size charts, composition. Ask in which format, who uploads it, and what happens with new products each season.
- How does it integrate with the site? A snippet, a platform app, a custom integration. Ask how much work your developer needs to do and whether the backend is touched.
- How long does it take to go live? Ask for an estimate for your catalogue, not an average.
How results are measured
- Can you run an A/B test? It is the only way to know how much the tool changes conversions and returns on your store. Ask whether the vendor supports it and on which metrics.
- Where do the numbers they show you come from? Return reduction percentages without source, period and method tell you nothing about your case. Ask how they were measured.
Data and contract
- Which shopper data is collected and where is it processed? Ask for the list of data, where it is stored, for how long and whether it is used for other purposes.
- How does pricing work? Flat fee, by traffic, by orders. Ask what happens in peak months.
SizeFlow's answers
- Data asked of the shopper: height, weight, age and gender.
- Calculation: a mathematical algorithm derives the shopper's body measurements and compares them with each garment's measurement chart, managed in the backend. It is not an artificial intelligence model.
- Integration: a snippet on the product page, without touching the store's backend.
- Measuring results: the effect is verified with an A/B test on your store.
For the other questions, from set-up time to pricing, the answer depends on your catalogue: we give it to you in a 30-minute demo on one of your products. To estimate how much size-related returns weigh today you can use the return cost calculator; module details are on the SizeFlow page.
Frequently asked questions
What is the most important question to ask?
How results are measured: without an A/B test on your store you do not know how much the tool changes conversions and returns.
Does a size recommendation tool need artificial intelligence?
No. It can rely on a mathematical algorithm, a trained model or both: what matters is knowing how it works and being able to explain it. SizeFlow uses a mathematical algorithm.
What data does SizeFlow ask the shopper for?
Height, weight, age and gender. From these it derives body measurements and compares them with each garment's measurement chart.