top of page

A.I Gift Generator Case Study

A.I assisted gift idea generation and occasion calendar with reminders. 

Screenshot 2023-07-17 at 5.14.05 AM.png
  • 50% of users report they would use A.I to help shop for personalized gift ideas.

  • 75% of users like the ability to thumb up and down search results.

  • 75% of users express the calendar and reminders to be a helpful feature.

Role: UX Designer        Project: Inspiration Idea      Duration: 4 Weeks      Tools:  Adobe XD Leonardo A.I | Mural | ChatGBT

The Problem

Users have difficulty and get stressed thinking of gift ideas.

The Objective

Assist users in remembering and finding personalized gift ideas.

Step 1: Empathize

I began with user research, creating a google forms survey and sharing it on the social media platform, Facebook.
User Research Goals:
  • How do users feel about giving gifts?
  • What are users financial impacts from gift giving?
  • How do users plan their gift giving? 
The insights I learned from the user research revealed that many users are stressed both mentally and financially by gift giving and often procrastinate doing it.
Screenshot 2023-07-02 at 10.56.57 AM.png
Pie Chart Showing 58% of users find big holidays financially stressful
User Research Insights:
  • 58% of users find holidays financially challenging.
  • 32% feel frustrated and stressed thinking of gift ideas.
  • 47% of users procrastinate buying gifts.
I used the insights from my user research to create a user persona.

If your company needs user research done, contact me.
Screenshot 2023-07-03 at 8.52.46 AM.png

Problem Statement

Buying gifts is often a mental and financial burden for users.

Step 2: Ideate

I brainstormed with Mural, incorporating user research responses from my affinity diagram.
I expounded upon the affinity diagram insights in the brainstorm to get more specific. I thought about the pros and cons of each category noting that the easiest is to just ask someone else which removes the mental burden. 

A.I is on demand, will never tire of being asked things or tell anyone you needed to. A.I could also combine several of these solutions to allow for multiple use cases. 

Step 3: Design

Since this project had an existing website, I marked up aspects I could iterate on.
Next I created a close up storyboard so I could visualize what pages needed to be made and create those wireframes. 
Screenshot 2023-07-17 at 1.13.15 PM.png
I wanted to personify Alexa as a face for Amazon's A.I so I used Leonardo A.I to help create a visual. Alexa currently presents as female and I wanted to represent diversity so I was very intentional with my prompt engineering. Regardless, I had to iterate a few times for good results.
Screenshot 2023-06-29 at 7.32.41 AM.png
Screenshot 2023-06-29 at 7.33.48 AM.png
Screenshot 2023-06-29 at 7.15.19 AM.png
Initially I thought about having users enter information through a form, but no one likes forms so then I thought of a series of individual questions to answer with a progress bar, but finally settled on a chat box because that is more conversational and I want the A.I experience to feel social.
Screenshot 2023-07-17 at 5.39.38 AM.png

User Research Clarification and dropped feature:

The first survey in user research showed a 58% of users get financially stressed during big holidays. My solution to this was to incorporate Amazon Reload into the design but I needed to see whether users already used it or not. So I put out another Google forms survey which showed that users didn't really know about it but more importantly, wouldn't use it as I thought. My plan was to encourage users to use Amazon Reload to save up for big holidays which would help solve their problem of getting financially stressed over the holiday but also secure more Amazon loyalty and business. This idea fell on it's face so I let it go and focused on just the A.I gift assistance feature. 
Screenshot 2023-07-17 at 4.59.52 AM.png
Screenshot 2023-07-17 at 5.00.34 AM.png
 I conducted a moderated usability study to track users click paths. Most users had an easy time completing both tasks with minor iterations to be made such as ensuring certain text was linked to where the user expected. This was an oversight on my part I am glad the 
usability test pointed out. A fun idea that was introduced in user testing was letting users have the ability to click and drag items to cart but I'm not sure how to make this in a prototype.

Accessibility Considerations

Staying consistent with Amazon's brand identity, I used their style guide hex color codes which meet color/contrast accessibility requirements according to
Screenshot 2023-09-01 at 10.18.06 AM.png
Screenshot 2023-09-01 at 10.25.21 AM.png
The Amazon Ember font is sans serif which makes for easier readability and all of the typography is at a minimum of 16px or larger.  Other accessible features are available such as:
W.C.A.G 2.1.1 - Keyboard Navigation
W.C.A.G 3.1.5 - Reading Level 
W.C.A.G 1.1.1 - Alternative Text for Images

Screenshot 2023-09-05 at 10.01.11 AM.png

75% of users report the calendar feature as helpful.

75% of users like the ability to thumb up and down search results.

50% of users would use A.I to help with gift ideas.

50% of users prefer this layout of finding friends' wish lists over the current way.

  • Tracking number of user engagement over the year. 

  • Connecting conversion rates from A.I's suggestions to cart checkout.

  • Follow and interpret user flow analytics.

  • Importance of dropping a feature that research or testing shows isn't wanted.

  • Users would like the ability to drag items to their cart.

  • Necessity to be very specific in prompt engineering to get desired results.

bottom of page