Sales associates know their products, but no one can remember everything. ElmAssist fills in the gaps by pulling up product details on the spot, turning hand-drawn sketches into floor plans, and suggesting pieces that complement what a customer already owns at home. Rather than letting a sale fall apart over a fabric that won't arrive in time, associates can confidently offer alternatives that fit the customer's home and their schedule.

Elm-Assist

Role

Tools

Designer

Figma, Figjam

The problem

West Elm sales associates must know hundreds of products with different materials, finishes, and fabric options while helping customers design their spaces. Customers often want specific combinations within a certain delivery timeframe, but those exact options are not always available. Without quick access to alternatives, customers may leave without purchasing.

When associates do not have the right information in the moment, it becomes harder to confidently recommend alternatives that match what the customer wants. This can lead to lost sales and missed opportunities to upsell additional pieces that would complete the room and lost sales means West Elm unnecessarily loses money. A sales enhancement tool could help associates make faster, more confident recommendations.

The Solution

An AI-powered tablet assistant that gives West Elm sales associates instant access to product details, finish and fabric availability within a customer’s delivery window, and personalized recommendations based on purchase history. Associates can also photograph hand-drawn sketches to instantly generate digital floor plans, making it easier to visualize spaces and recommend products that fit the customer’s needs. By helping associates find the right products faster and suggest better alternatives, the tool creates a smoother, more personalized experience for customers and the opportunity to upsell and increase sales.

Talking to an expert

My interview with a Design Crew Member

I interviewed a Design Crew member at a West Elm in the Philadelphia suburbs to find out their pain points when it comes to selling furniture and making floor plans for clients. She gave me insights on two types of West Elm employees that struggle with information overload:

Self-Guided Research

I was curious about the broader impact sales associates had on buying power. I referenced McKinsey, Salesforce's State of the Connected Customer report, and the National Retail Federation for data.

    • 73% of customers say a good experience is a key driver in their purchasing decisions

    • 86% of buyers are willing to pay more for a great customer experience

    • Positive in-store consultations lead to on average of 30% more spending than those who browse independently

    • 60% of furniture shoppers say they feel overwhelmed by the number of options available

    • Customers who use design consultation services have a significantly higher average order total than walk-in customers

    • 40% of furniture purchases involve at least one in-store associate interaction before the final decision is made

    • AI-powered retail tools are projected to generate over $340 billion in value for the retail industry by 2028

    • 69% of retailers say AI has improved their ability to personalize the customer experience

Key Insights

  • Every unanswered product question is a missed sales opportunity, which means West Elm loses money

  • The difference between a lost sale and a closed one often comes down to what the associate knows in the moment

  • Customers who feel confident in their choices spend more and return to make more purchases

The case for empowering sales associates is established. The data shows that better associate knowledge doesn’t just improve the customer experience, it drives sales.

What My Research Taught Me

My research confirmed that sales associates are the single most important factor in driving sales - so much so that you can even talk to an in-store associate while purchasing online (a feature called Design Chat.) West Elm’s retail experience is built around knowledgable sales associates, but there is an overwhelming amount of information for them to remember.

Wireframes

I used this process to expand on how the experience of searching for information would come together. I decided what the homepage should be (the general search) since it would be used the most. This process also helped me keep in mind how a user would navigate through the different screens and potential user flows that would come about while using ElmAssist.

Designs

The general search feature’s AI capabilities are highlighted here so any associate, new or returning, is reminded of it what it can help them with.

Search/Homepage

The general chat interface is clean and simple, designed to blend in with West Elm’s current branding. It can handle any type of question from the search bar on the bottom of the screen.

General Chat

The product availability feature is designed to give associates real-time info on products within certain client constraints. In this example, a client wants a certain type of sofa by a certain date. ElmAssist will only show the fabrics and finishes this sofa is available in by that time, avoiding the possibility they choose a fabric that isn’t in stock by their project date.

Availability Chat

The customer search feature surfaces the associate’s 5 most recent clients for quick, easy access. There is also an option to search for clients by their name, email, phone number or order number.

Customer Search

Design Crew, West Elm's current digital floor planner can take hours to develop custom floor plans for multiple rooms. ElmAssist eliminates that entirely by using AI to instantly transform hand-drawn sketches into floor plans, automatically including products it knows will fit the space. Its AI then factors in client deadlines, filtering recommendations in real time to only show products that can be delivered on time. It also takes into account the customer’s purchase history, recommending cohesive pieces for the space.

Sketch to Floor Plan Chat

Test and iterate

Once I had my designs fleshed out, I wanted to understand how useful they would actually be in real life and how I could add value to them. I showed the main screens to the same Design Crew member I interviewed for my research, asking a mix of the following questions for each screen:

  • Looking at this screen, what are your first impressions?

  • Do you have any questions?

  • How often do you think you’d use this?

  • What else would you want to see that would make this more valuable?

She said she understood what I was trying to do, but asked for another feature. Based on her answers I created an additional screen based on what would make the product more valuable and easier to use. Overall, the feedback was very helpful and gave me clearer idea on what was working and what needed more thought.

Availability Screen: Client View

One of the features she noticed and spoke more about was in the availability chat, where there could be an option to show the customer in real time what fabrics and finishes are available on a certain piece by a certain date. She said she wishes she had something like that, instead of having to click through multiple options on the computer with the client where they might see something they like, only to be disappointed it wasn’t available for 10+ weeks.

What i learned

I learned the importance of designing with empathy for users who can't always articulate their pain points. Retail associates are experts at their jobs, not at articulating UX problems. Therefore, it was crucial to research and observe workflows. From that I could translate their frustrations into concrete design opportunities. This taught me that good UX research isn't just about asking the right questions, but about knowing how to interpret the answers you get. Users can tell you what frustrates them, but rarely how or why it should be fixed.