Zappos New Feature Idea 2: Fit Recommendation

Background

This is an exercise to develop a new feature idea for Zappos. Below is the steps I took to get to this idea, the feature itself, and a potential way to implement it.

User Research

Population & Technique:
I interviewed some online shoppers and zappos users.

Question type:
Their previous shopping experiences.
How they make a judgement in terms of which item might fit them.

Insights:
Some users choose sizes by comparing the item they want to purchase with their previous shopping experience of the same brand. However, this measurement is not always accurate since different model may be sized differently even within the same brand.

Zappos Feature Inventory

I inventoried Zappos' current features so that the new feature could be build upon the exisitng system.

The Idea

Based on the steps above, I create the following idea.

We'll see how this feature works through a Zappos user Taylor. On a tuesday afternoon, Taylor decides to go onto Zappos and shop for a new pair of shoes because he loves Zappos and browsing on Zappos is one of his leisure activities. Besides, he wants to add a new pair of black converse to his shoe collection so that his new converse collection could include the full color spectrum. When he looks at this item, a pair of black converse, Zappos suggests that size 8 might be the best fit for him

The fit recommendation feature makes this suggestion and it's based on Taylor’s previous shopping history. This recommendation feature predicts Taylor’s size range by looking at the measurements of previous shoes purchased. Taylor can make this sizing model more accurate by rating how his previously ordered items (including returned items) fit him. This model would adjust itself based on additional feedback from Taylor. Moreover, Taylor can select which items to be included in his sizing model. If Taylor has never purchased an item at Zappos before, he could give this recommendation feature some data by finding and rating several products that he already owns. Now Taylor has some good suggestions and he feels more confident with the size of the converse that he is going to buy.

The mockups for various devices:

pop-up explaining this feature for users

android

ipad

How this could be realized:

This feature is a combination of Zappos' current Fit Survey and order history feature. First, Zappos would need to associate user’s fit survey result with this feature. When calculating the sizing model, Zappos would use the order history feature to get a list of previously purchased item. What needs to create is a database of the shoe size information. This information probably could be found via Zappos’ brand contacts. Moreover, to enable this feature, we need to develop an algorithm that could calculate and identify size ranges based on the items purchased. This algorithm should able to identify relevant information and could categorize itself. There should not be too much change with the UI besides adding some additional components and requesting users to fill out a fit survey for every item they buy and every item they return.

Main Page