The app offers 3 different types of recommender. How do you decide which one is best for you? Here we give a brief overview of each type to help you decide.

First, what is a Size Recommender?

Size Recommender is the new (better) way to help shoppers figure out their right sizings. For a very long time, all online shops use size chart tables but these tables are not inherently user-friendly and poses a big friction for online shopping. This is an even bigger problem if you operate a shop with sizes from different vendors.

Size Recommender solves the problem by asking the user to enter their measurement once, and then recommend the best fit. Then on every subsequent product pages, we will automatically recommend the size for these products using the previously given information.

1. Advanced Apparel Recommender:

  • What is this?
    • User enters basic information (i.e. age, weight, height) and we estimate the user's body size which is then used to make apparel size recommendation.
  • Requirements:
    1. Currently only works for Adult Apparel (Men & Women)
    2. Need a sizing table. The measurements also need to be linked to canonical measurements.
    3. Size chart need to set up product info (I.e. product type, age, gender)
  • Pro:
    • Low to no friction! Not everyone knows what their chest or waist measurements are but almost everyone knows how tall and heavy they are. This type of recommendation will have the lowest drop-out rate which will ultimately improve your conversion. 
  • Cons:
    • Currently offered only in English, French, Spanish, German, Portuguese, Italian, Japanese, and Chinese
    • Don't work for certain products types like pet, toddler, and maternal clothing. We can currently only estimate body sizes of male and female adults. 
    • Needs more information from the size chart (i.e. canonical measurement linking, product info) to work

2. Simple Generic Table Apparel Recommender:

  • What is this?
    • User input different body measurements, the app makes the best size recommendation based on that.
  • Requirements:
    • Only works for apparel based products (i.e. doesn't work for shoes 
    • Need a sizing table
  • Pro:
    • Any apparel products with a size table can use this recommender type
  • Cons:
    • Higher friction for user: not everyone knows their body measurements

3. Custom Recommender

  • What is this?
    • A programmable recommender where you can design the entire recommendation logic. Examples of product that is suitable for custom recommender includes bras. 
  • Requirements:
    • You need to know the exact logic for recommendation
  • Pros
    • Extremely flexible because you have complete control over the logic. So you can ask for arbitrary input.
    • It can potentially be used for things outside of size recommendation. 
  • Cons
    • Takes a lot more work to set up. There is also a learning curve to understand how the component works.