UX research, UX/UI design, strategy
The Home Depot Job Materials List Tool: Streamlining customer product conversion process to win the job
overview, essentially
Enhancing competitiveness through efficient job materials list transformation 🛠️
In a rapidly evolving market, The Home Depot (THD) has identified a critical challenge that necessitates an innovative solution. The existing procedure employed by THD Associates to transform a job materials list (JML) into a comprehensive list of THD products and SKUs has been recognized as cumbersome and time-consuming. This inefficiency not only hampers internal operations but also impacts THD's competitiveness in comparison to industry peers who offer streamlined and automated methods for obtaining quotes based on job materials lists.
My team and I were asked to craft a user-friendly solution that redefined how THD Associates transform customer job materials lists into comprehensive THD product lists. The ultimate goal was to elevate THD's market presence by providing a quicker, more intuitive method to generate precise quotes from customer job materials lists. For this project, I was the designer driving to not only meet industry standards but to set new pinnacles for customer-centric service and operational leadership.
Project Timeline: 3 months
Role: Led the research, concept creation, interaction design, and prototype testing for MVP experience
Team: strategist, product manager, dev team, product designer (me)
Partners: The Home Depot
process
Utilizing the Hypothesis to Action (H2A) framework
The team went about our sprints utilizing a Hypothesis to Action (H2A) framework to help drive the research process and prioritize backlogs in order to maximize gains.
At the beginning of each sprint, we outlined 2-3 major hypothesis to test within sprint, either through testing with users, conversations with key stakeholders, or feasibility checks with THD developers.
After research had been done on each hypothesis, we would synthesize our findings, distill convictions from them, iterate on design/experience work, and finalize immediate actions to be done in the next sprint.
Board of stickies to help organize research findings each sprint
Example of our H2A board to keep track of major findings, convictions, and actions throughout sprint
PERSONAS
Both personas are referred to as “THD Associates” throughout the case study.
Pro Account Rep (PAR)
“I want to partner with my customers to provide solution sto all of their projects and grow both of our businesses”
Key insights:
Wants to be in partnership with their customers
Wants to build and maintain relationships with customers
Thinks that THD has limited systems enabled to help PARs win the job
Pro Account Sales Associate (PASA)
“I need streamlined systems and tools that help me service my managed customers”
Key insights:
Highly skilled at solving complex customer requests
Frequently short-handed
Challenged by slow systems and broken tools at THD
Needs reliable tools and technology to improve customer turnaround times
first step
Understanding key pain points in current state translation journey
We conducted interviews with different Associates to understand what their current process looked like for translating a generic JML into a THD specific quote.
Spoiler: it was laborious, outdated, and overall a varied process from person to person.
We wanted to understand what types of formats and inputs they received, what their process was for searching and matching items based on data and type of use case, how much they engaged with the Pro when finalizing a list, and generally what opportunities might exist to improve the current process.
Key Pain Points in Existing State Journey
Time-consuming, multi-step process that varies from person to person, team to team
Variabiltiy of JMLs received makes it hard to streamline
Manually looking up THD products/SKUs based on a Pro given JML through various channels (e.g., THD channels and 3rd party channels)
Once found, manual entry of THD SKUS into THD quoting application
Balancing feasibility and desirability for MVP
Through this research, we found that our personas receive a wide variety of materials lists from their customers (e.g., handwritten notes, Excel sheets, list in email).
By regularly speaking with our engineering team members, we knew we had to make a choice on what type of input our MVP tool would handle first. We learned from the engineers that the more digital and structured a materials list is, the more feasible it will be ingested into a digital translation tool.
Based on this and our research on what would be highly desirable from users, we decided to narrow in the focus of our MVP to intake materials lists that could either be copy/pasted in or uploaded via a spreadsheet.
Examples of wide range of Job Materials Lists received from Pros
Understanding what main features/functionality the MVP should have
I conducted users tests to find out what our users desired path was for the input portion of our tool
What type of information must we require for tracking purposes?
What type of materials list information can they enter?
How do we allow them to enter their materials list (e.g., copy/paste, upload, or both)?
How much work should the user do upfront before coming to our tool? How much are they comfortable doing?
What experience improves speed to quoting?
I then spoke to users on what their desired path would be for the output portion of our tool, the translated list
How much work are the Associates willing to do to get to a “complete” translated list?
How accurate must the list be upon first translation for the user to feel it’s “worth it?”
Would users rather see a partially translated list first, and then work with the tool to narrow down on a more accurate list, or would they rather do the narrowing work first and then see their full translated list?
Input must-haves for MVP
Our tool must pursue a hybrid approach to include both form field inputs AND upload functionality.
Output must-haves for MVP
Our tool must clearly produce increasingly more complete / accurate / confident results in order for Associates to be willing to engage
MVP tool must produce a partial list of results first before allowing user to narrow down results
Our tool must be able to allow Associates to easily engage Pros in the process, if desired, and connect to a broader quoting experience
feasibility check
Aligning with engineering to connect backend systems to frontend
user testing
Initial design user testing
Knowing that the MVP will consist of an “input” and an “output” portion, I conducted user tests with actual THD Associates via Zoom to understand what flow would be more preferable to them.
Input Option A: take the user through a series of questions answering them one by one to get information, required more page loads
Input Option B: single page view input of information
Output Option A: More info at first glance, dropdown for details, circular progress bar
Output Option B: Product description only at first glance, right panel for more detail, & line progress bar
Final prototyping
Getting into detailed interaction design and user testing full experience
Based on the user testing sessions, we now had a clear path for our MVP and it was time to refine the interactions and work out any edge cases that had come up along the way.
At this point, we were ready to move into detailed design patterns but quickly learned that THD did not have a consumable design system in place. We had to work closely with our dedicated developer to understand what components were already built and what would have to be net new.
Finding a better match if the tool doesn’t get it right the first time
The user will be able to see a “match score” for each line item based on the level of confidence the tool has given the information that was inputted.
When clicking the dropdown to see more details for an item, a user can decide if the item matches what the Pro was looking for, and if not, they can click in to narrow down search results further to get a better match.
Exporting translated items to engage the Pro for review
If the Associate has questions about a line item and wants the Pro to review it, they can directly export that specific line item to send to the Pro. Users can also export the full list as either a PDF or Excel file to send to Pros or move directly into a broader quoting tool with.
Need to scale back MVP experience to meet targeted build time
With appetite for this tool to be built as soon as possible, a goal build time of 12 weeks was set in place to pilot the MVP with a certain group of Associates.
Once we finally onboarded a dedicated developer to the team, we learned that we had to scale back some parts of the experience to meet that 12 week build time. This required us to redesign a part of the experience and run quick tests to validate that we were still solving the problem for our users.
Learnings
Although we were engaging the engineering team throughout the project, the team and I took this as a learning to advocate for a dedicated developer on the team as early as Sprint 1 to make sure that business stakeholders have a clear voice from engineering early on in terms of feasibility to avoid any scale back of MVP later on in the project.
Keeping key stakeholders engaged throughout the way
Each week, we met with major client stakeholders across product, engineering, and design to update them of progress, key decisions, prototype updates, and ask them any major questions relevant to the MVP experience.
I also made sure to get key stakeholders excited about the designs we had been creating by bringing them along in the design process and highlighting users excitement about the tool we were creating.
Outcomes
Final delivery and outcomes
The tool is currently being built by Home Depot internally and the MVP will be launched in pilot to test, learn, and iterate.
Over time, increasing quoting workflow efficiency for Associates will increase incremental annual benefits of $115-$233M and will overall improve the Pro’s experience and view of Home Depot as a viable partner for their business.