Arnold is a smart sales support system that delivers in-the-moment guidance to retail sales teams.



SMC IxD Senior Studio




A.J. Mody
Ana Massette
Evan Pierre
Orhan Basak

MY Roles

Project lead
Design research
Interaction design



Find use cases for IoT in apparel retail.

In Spring of 2018, SAP came to our senior studio class with a design challenge: How might we help retailers to convince shoppers to buy more in the store? The project objective was to find new use cases for the Internet of Things (IoT) in apparel retail. Our solutions had to be research-derived and involve both hardware and software.


Smart, real-time sales support.

Over 16 weeks, my team and I developed Arnold, a system that makes IoT data useful during operations. Using artificial intelligence, Arnold takes IoT sensor data, combines it with existing sales data, and makes suggestions that help employees make more sales.


System Overview



Challenging the brief.

Our first step was to challenge the brief. High-level desk research helped us get familiar with the retail space and set context for our project. We found that employee satisfaction is directly connected to customer satisfaction, which hugely impacts overall company performance. Recognizing these links, we reframed the original HMW into: How might we improve the employee experience?


Understanding the space.

To gain a holistic understanding of the current retail environment, we conducted field research and one-on-one interviews with sales associates, customers, and managers. Based on our findings, we modeled three personas.


PERSONA Snapshots



Pressured store managers often become controlling. This behavior creates fear and frustration in their sales team.

When looking at our primary findings, we came across an interesting collision. Store managers are often stretched thin and start to micromanage out of fear of losing control over their team and operations. However, many of the sales associates that we spoke with expressed high levels of dissatisfaction with this type of behavior. Rather than feeling supported by their managers, they feel controlled, and many of them are afraid of making mistakes or asking for help.


Sales associate in response to "What do you dread about work?":

"Getting [my manager] mad... Just that."


Connecting EX to sales.

Based on above insight, we formed an assumption: If we use technology to lift responsibilities off the manager's shoulders, she will be better able to support her employees, making their lives and work more pleasant. But how do we translate an improved employee experience to increased sales? To identify areas of opportunity linked to the business bottom line, we invited a store manager into our design studio to co-create a journey map of a typical day with us.


Journey map Co-Creation



Turning pits into peaks.

The journey map helped us identify two pretty awful windows, both representing peak sales hours. During peak hours, the sales team has the most amount of customers in the store and therefore they can make the most sales. However, these times are the most chaotic: customers are rushed, sales associates are overwhelmed, and the manager needs to make sure that operations run smoothly. We decided to focus our solution on this time of day and conceptualized an intelligent assistant manager, Arnold. Using existing sales data, IoT, and AI, Arnold helps the entire team make the most out of these busy but lucrative hours.


Use case scenario



Delivering useful support in the heat of the moment.

To define what type of support would be useful to a sales team and learn when and how it should be delivered, we tested a low-fidelity prototype with a store manager. During the session, we learned that Arnold should leverage the walkie-talkie system as text-based delivery turned out to be complicated.




Final deliverable

How Arnold works.

At the end of the project, we presented our concept and video to SAP. Based on insights from usability testing, we decided to focus our story on the Arnold voice interaction.