GV Design Sprint for an AI-enabled interior decorating tool. Spring 2019.

The Problem: Moving into a new apartment or home is a stressful event that everyone must encounter a few times in their lives. Decorating that new apartment or house and turning it into a home is an even more stressful event that requires a lot of time and money. In our society today, time and money are both luxuries people cannot afford to waste. House2Home is an e-commerce site for home decor and accessories offering starter kits for customers to confidently personalize their new space seamlessly without wasting time looking for specific pieces.

The Solution: A modified GV design-sprint approach lasting 5 days focused on mapping, sketching, deciding, prototyping, and testing.

My Role: GV design-sprints are usually completed in teams, however, I challenged myself to complete the sprint in a team of 1. I reviewed customer research, as well as created sketches and created a mid-fidelity prototype that was tested on 5 different customers.

Tools Markers, sketchpad, sticky notes, Figma, InVision, Zoom, Google Hangouts, Skype, and a phone.

Goal: Create a starter kit with items ranging from $10-$50 for new living space for the website only e-commerce site House2Home.

Day 1 of Sprint

Discovery

The user customer research given to me consisted of survey results stating that “customers who just moved to a new home or apartment want to personalize their new space, but don’t feel confident to decorate.” I also read through several personas like Lindsay who is budget conscious or Renata who feels inspired but doesn’t know what to buy. After reading the brief, personas, and research, I began formulating ideas.

Day 2 of Sprint

Ideas

I began thinking of possible end-to-end UX for the customer.

Day 2 of the sprint brought on more sketching and ideation. I looked at several products for design inspiration including Instagram, Spotify, Wayfair, Amazon, and a UK based product called Uni Kit Out.

Amazon’s Home Style Quiz gave me major inspiration for the quiz portion of this design.

Another much lesser known product that inspired me was Uni Kit Out. This is a UK site that helps new students create a kit for either their bedroom, bathroom, or kitchen. The customer gets the kit delivered to them the first day they arrive at their new accommodation thus eliminating the customer arriving to an empty space.

After feeling properly inspired, I took Sharpie to paper.

To start off my brainstorming session, I did the Crazy 8s exercise. I decided to go with the first screen I drew as the critical screen because I knew I wanted to base the kit results from a quiz like Amazon’s Style Quiz. In my mind, the customer needs to take the quiz before seeing the kits.

Day 3 of Sprint

Decisions

I decided to stick with my initial Solution Sketch and build from there because I wanted to implement artificial intelligence via the quiz. I am personally fascinated by new emerging technologies, especially artificial intelligence, and felt this product would greatly benefit from the implementation of an AI. I believe this is what customers want when they look at House2Home’s Decor Starter Kits. Customers don’t have the time to shop for individual pieces, most of them don’t have the budget for it, and the majority doesn’t feel confident to stick to a theme and decorate their new space.

Day 4 of Sprint

Prototype 

I made the prototype using InVision. I understand this is a basic prototype from a visual standpoint, so I’m excited to see how testers react to it on Day 5.

After clicking the create button, the customer will be prompted to a quiz. The quiz consists of several questions asking the customer to choose between styles and decorations. Results will suggest which kit is for them.

Artificial intelligence meets HGTV.

After the quiz, the customer will see the Results page with the best matched kit showcased up front. The customer can then personalize their kit by choosing which room in the house, which kit (basics or deluxe?), and color combination. If the customer isn’t happy with their match, they will see two other matches on the same screen as well. If dissatisfied or curious to see other matches, the customer can redo the quiz or if they are satisfied, take the next step and make a purchase.

With this prototype, I hope to learn the following: 1 How well people respond to AI serving as their interior decorator 2 Usability and ease of flow 3 Feedback to improve future iterations.

Day 5 of Sprint

Testing

I interviewed a combination of both males and females between the ages of 25 and 35 with an interest in beautifying their homes. All tests were done remotely via Google Hangouts and Skype.

The tasks given to the testers were as follows: 1 First impressions of kit page 2 Take (or pretend to) the quiz 3 Secure the bag 4 Place Order

All 5 testers said the flow from start of quiz to confirmation was easy and straightforward. One tester used the adjectives “accessible” and “intuitive”  to describe their overall experience. Interestingly enough, 3 / 5 testers mentioned the promotion code in the check out screen. They all said they would firstly check online for a promo code to see if they could get an even better deal.

Two out of 5 testers said they would use the product today if this were a real product. One tester said “I would use this right now if this was real. Nobody has time to look for decorations and stick to a theme, at least I don’t!” Most customers thought taking a quiz was fun and easy.

Critiques of the prototype were font use of logo. One tester said the logo looked “cheap” because of the font used in the logo. Other critiques were that they would have loved to see photos of each item in the kit. Overall, it was a quick and easy process despite being a mid-fidelity prototype. 

Reflections

Creating a starter kit with items between $10-$50 that met customers’ tastes and lifestyles was this project’s biggest challenge, which led to a creative exploration of home decor meets artificial intelligence. User research and sketching led to a mid-fidelity prototype that tested the concept of AI giving interior decorating advice when customers found themselves without time in their busy schedules nor the funds to buy select pieces. The prototype was well-received as testers asked when the real product was, if ever, launching.