The Results

HYPOTHESIS: Find product faster and more accurately

I believe that making it easier and faster for a user to find the product they are looking for will increase engagement and potential sales.

I will know this is true when I see a significant increase of more than 20% of users be able to find the product they are looking for in under 30 seconds.

The three steps below helped me formulate the main hypothesis that drove the rest of my usability experimentation.

I interviewed 5 current Amazon.com users and asked them some basic questions: 

• what do you use Amazon for?
• why do you like using Amazon?

And the major consensus answers were listed below:

• 5 out of 5 people said that:
  it has reviews!
• 4 out of 5 people said that:
  Amazon has everything!
  it’s so easy to find what I need
• 3 out of 5 people said that:
  there’s no need to go elsewhere to find information

Find the best value vacuum cleaner

Imagine that your vacuum cleaner just broke and you are in the market for a new vacuum cleaner. You are interested in purchasing the best vacuum you can buy on a budget. Your friend told you about a new thing called ‘Jet’ that has really good deals, show me what would you do to find it.

Here is the number:

• 1 out of 7 have heard of Jet
• 3 out of 7 would rather use Jet as a starting point, then go to Google or Amazon to read reviews & potentially buy
• 7 out of 7 experienced issues surrounding product information page
• 4 out of 7 users were unable to immediately find what they were looking for
• 6 out of 7 experienced searching issues

Ultimately, users were not able to find what they were looking for!

Affinity map helps me decide on which pain points to address first. Although 7 out of 7 users were looking for product reviews, I had to prioritize the critical issues first. In this particular case, users will not even arrive at the product review problem if they are unable to find the product that they are looking for.

Therefore, the most important issue to address is: searchability

Uncover design opportunities through user flows

Below is the modified user flow. The yellow lines represent potential design opportunities that I further explored during my ideation step.

After sketching several ideas, I spent some time going back to the user flow and using dot stickers, I identified which ideas can potentially solve which pain points.

The winning solutions were the ones that solve more than 1 issues at a time.

These selected solutions were then brought into high-fidelity prototypes and went through several rounds of iterations before arriving at my final design recommendations.

1. Make the search bar more prominent by treating it as the main call to action on the Home Page

2. Display recently searched and viewed products on home page

Here is how I imagine the interaction of the new home page will be:

3. Smart default sorting on Category Page: alphabetically

• future opportunity: learn users’ interest over time and push categories that are more relevant to them to the top of the page

4. Display search icon on Category Page

• provide a quick way for users to go back into ‘searching’ mode if they are unable to find the relevant category of the product they are looking for

5. Display past and recent search queries used on empty searching state
(when user tap on search to begin searching)

6. Display both categories and sub-categories on search results

• provide a quick way to help user refine their search (for example: shoes under fitness category will inherently yield different results compared to shoes under clothing category)

7. Use Predictive Text to quickly provide suggestions to users

8. Utilize ‘empty searching state’ page by displaying past/recent search key terms

9. Provide a quick way to help understand why they land on this particular page

10. Provide reasonable suggestions based on the current search query used

11. Enhance sorting and filtering interaction

Here is how I imagine the interaction of the new filter and sorting feature will be:

I decided to create 2 different clickable prototypes in order to display both conditional situation that might arise during user’s process of finding a specific product.

One Click Test

I took the ‘Success Search Flow’ prototype out into the wild and ran a One Click Test using the same scenario from the initial guerrilla usability test.

I approached 7 users and told them the same scenario.

Based on a one-click test, 7 out of 7 users would go directly to the search bar in order to accomplish their task (finding a vacuum cleaner)

This is a greater than 40% increase in the chance of users’ ability to find the product they are looking for—compared to the initial guerrilla usability test result (increase from 54% to 100%)

By the time I presented this project to a PM from JET, the iOS application has gone through so many rounds of iterations and therefore no longer look like the version that I worked on! Even so, through my conversation with Lizzy Kiat (a PM at JET), I learned that there were plenty of internal constraints in play when JET launched their first couple versions of iOS app.

With new information on hand, the next reasonable steps is to run a similar test on the new version of the app and see if similar issues still persists.

Working on JET iOS Application taught me to never be afraid to reach out to someone from the company for the product that you are analyzing! Not only that I was able to present my findings to real stakeholders, I was also able to learn more internal constraints and understand the reasonings behind why specific design decisions were made.