Nearly $50 billion was lost last year to retail theft in the US. As an intern, I designed a tool to help The Home Depot unmask these criminals. 

My Role    

  • Led the end to end UX process

  • UX Research: conducted observations,  interviews and user testing. Completed the affinity mapping process and designed information architecture

  • UX Design: prototyped experiences at low and high fidelity through concept sketches, wireframes, and interactive prototypes. Redesigned the system utilizing THD's existing design system

  • Co-facilitated an all day offsite workshop that led 70 of our H.R. business partners in user persona creation and affinity mapping

Timeline    

May 2019 - August 2019

Primary Prototyping Tool

Sketch

Intern Team​ Members

Daniel and Naasom

Results​

  • Aided in enhancing a system used to catch criminals stealing from The Home Depot store​s with a projected initial decrease in retail theft of $400,000.

  • Reduced the time users spent navigating between apps and calling store managers which is projected to result in saving users 10 minutes per use in an application used mutliple times a day.

  • Enabled business partners to empathize with the journey of potential applicants resulting in their desire to improve the on-boarding process for thousands of applicants.

To comply with my non-disclosure agreement, I have omitted confidential information in this case study.

Background

15 years ago Tandem Web Tools was created as a series of search tools used to pull information about sales.

This suite of tools has been built up over the years and includes search tools for gift cards, receipts, deposits, refunds, and more. 

 The Problem

The user interface of Tandem Web Tools did not match the company's current design system, it had a very complex information architecture, and the necessary functions that users needed to do their jobs were missing.

 Process

I have no idea what this tool is, but I would use it if I knew it was there. It could be so helpful.

Store Associate

 Observations & Interviews

Originally we couldn't find anyone who had even heard of the tool. After asking many people and being very persistent, we finally found our users. Next, we began to conduct semi-structured individual and group interviews with each of our users.

We chose to observe users while working as they utilized Tandem Web Tools in order to better understand how the tool is used. We also interviewed users in order to learn more information about their history of use with the tool, as well as their challenges, wishes, and favorite parts of the tool. 

24

participants interviewed

3

seniority levels represented

departments represented

Affinity Map

After interviews, we chose to complete Affinity Maps in order to organize the information we learned from each user group.  

We decided to create separate affinity maps for each user group, because many of the user groups were using different functions within the tool. We wanted to easily separate each group's unique pain points.

From our interviews we learned about a new tool called Magellan. 

This tool was intended to replace Tandem Web Tool 4 years ago. However, this new tool was a thick client tool, was not accessible to all of our users, and did not follow Home Depot's current design system. 

We used feedback from our users to present the idea of creating a new system that combined and updated the functionality of both tools to our team. Our team agreed and we began to move in a new direction.

 Users & Pain Points

From our interviews and affinity maps we identified  many pain points and design ides provided by our users. A few major pain points for each of our users is presented below. 

Below I identified a few of the major pain points for each user group.

Financial Analyst

  1. Doesn't understand the logic behind reports and codes

With all of the information gathered, user personas were then created. Unfortunately, I can not show the User Personas that were created. 

Feature Prioritization

Our users gave us many ideas for features they would like to see when we interviewed them. As a team, we discussed all user pain points and feature requests through the lens of our users. We utilized a Prioritization Matrix in order to decide which features and user pains should be addressed in the first iteration of the design.  We used difficulty and value as the metrics for our matrix. 

One of the values of the Home Depot is "Creating Shareholder Value". We wanted to ensure that any new changes were aligned with that value. Also, our team wanted to be sure that we could focus on features that we would finish within the scope of our 3 month internship. Any features that were out of scope were added to the future recommendations we gave to our team at the end of the internship. 

 Problems to Solve

Our team Feature Prioritization session led us to identifying the following problems to solve during our internship:

How might we reduce the number of screens users have to navigate through to find information?

How might we make the logic behind reports and codes more clear?

How might we provide our users with one tool that they all can access and use efficiently? 

Ideate

Next, I created an information architecture diagram to help the team understand the navigation of the new application. As a team, we discussed any changes that should be made to the information architecture.  

Then, I began to generate sketches based on the "how might we" questions we generated, and all that we learned from our users. 

I got feedback on these sketches from our users and my team. Ultimately, we decided to move forward with the design that utilized many key components from the Home Depot's current design system. 

Our users gave feedback on these ideas, and they also liked the design that included navigation that was most similar to other Home Depot tools that they used.

Finally, Wireframes and Hi - Fidelity Prototypes were created to test with our users.

New Features

01

Tools were combined from Magellan and Tandem Web Tools to ensure that our users had access to all of the functionality needed to efficiently do their jobs. Users were also given the ability to see different views within the same page in order to reduce the need to navigate to different pages.

02

An information icon with a pop up help card was added to help our users make sense of reports and codes that appeared on the page.

03

We wanted to ensure our application was available to all of our users by making it a web based application that could be accessed through a URL. 

 User Testing

We put our prototype in front of our users to get their feedback on the first iteration of the design. We knew how valuable feedback from our users would be in determining our next iteration. Our users thought aloud while walking through the prototype, and we took notes in order to help us identify user actions, comments, confusion, and common themes. ​

After the user testing session, we discussed what we learned and planned our next steps. 

After completing user testing, we received the following user feedback:

Some users needed the ability to also search by drivers license in order to complete one of many key tasks within their role. 

Some users did not notice the help icon that was now available to them to explain the unlabeled codes within the application.

Tabs were created to offer users choice in the way that they could search, but many users did not notice the tabs.

Next Steps

Design, Feedback, Iterate

We need to utilize feedback from the user testing session to complete iteration two. Then, we should test iteration two with our users and get feedback. We would try to repeat this as many times as we could in to get as much feedback from our users as possible. This helps us to make sure we are truly meeting our user's needs.

Eliminate Tools 

Data was pulled by our software engineers to show that a few of the search tools had not been used at all this year. More research needs to be done in order to find out if there are truly any users of those tools. Tools that are no longer in use will be pulled from the application. 

Lessons Learned

Leverage Low - Fidelity Prototypes

It's better to come up with Low - Fidelity prototypes to test before moving to High-Fidelity prototypes. Low - FIdelity prototypes are fast and easily disposable.

 

Utilizing Low - Fidelity prototypes can save a company money in the long run. If you get too invested in the High-Fidelity prototype you mocked up and it is way off course, you now have to spend months going back to the drawing board.

Fail Fast

This experience taught me the "skateboard vs. car" ideology.

 

It is always better to design quickly and fail fast. It is helpful to get users something better that they can use to do their jobs as quickly as possible. The more iterations and user testing you can do the more informed your final design will be. 

Continually Involve the User

At every stage, it really helps to reach out to users to get feedback on the usability of your current design. This saves the company money and time.

 

It also helps ensure that you are truly meeting the user's needs.