Error Management
Error recovery for The Home Depot Price and Cost Management Merchants.
MY ROLE
UX Lead for Merch
TEAM
Talaia Whisenant
DURATION
3 Months
OVERVIEW
Home Depot Merchants use price and cost management software to adjust the prices of merchandise in the store.
During my time at The Home Depot I worked on the merchandising and pricing team responsible for creating price and cost management software. Merchants used this software to adjust the prices of merchandise in the store. In order to compete in the market price adjustments had to be made with the least friction possible.
I worked on helping merchants quickly recover from errors with clear messaging.
The worldwide information security market is forecast to reach $170.4 billion in 2022.
Gartner Research, 2018
PROBLEM
Merchants don’t have in-app access to their applications.
Productivity decreased for Merchant Planners who encountered unrecoverable errors while using the software to make retail and cost changes.
SOLUTION
Convert merchant accounts to single sign on with Connect.
Redesign the top errors that occur in the system in a way that resonates with Merchant Planners, and helps them recover from errors.
This will increase productivity and add to competitiveness in the home improvement market.
HYPOTHESIS
Consistent application branding and even clearer messaging will reduce friction in the flows.
I then conducted an assumptions and hypothesis workshop with my team to express all of our assumptions regarding error messaging in the software, and develop a testable hypothesis for starting research. I created a research plan to document our progress.
CHALLENGES
Unfortunately, the burden of SSO account creation falls on the user.
Alining the product team on this initiative.
Understanding software error messaging and error handling… Meeting with merchants. Getting a big enough sample size….
DESIGN PROCESS
The product teams developed 21 use cases and user flows to address the scenarios that could occur during after implementation. The main use cases of concern were new and existing customers creating new Connect accounts. We worked with UX writers to create clear single sign on messaging throughout the flows.
PRODUCT GOALS
The product teams developed 21 use cases and user flows to address the scenarios that could occur during after implementation. The main use cases of concern were new and existing customers creating new Connect accounts. We worked with UX writers to create clear single sign on messaging throughout the flows.
SCOPING
Error messaging best practicies.
I first researched the best practcoes for error messages and presented my findings to the product team to bring alignment.
PERSONA
The product teams developed 21 use cases and user flows to address the scenarios that could occur during after implementation. The main use cases of concern were new and existing customers creating new Connect accounts.
We worked with UX writers to create clear single sign on messaging throughout the flows.
Whatever it is, the way you tell your story online can make all the difference.
JOURNEY MAP
The product teams developed 21 use cases and user flows to address the scenarios that could occur during after implementation. The main use cases of concern were new and existing customers creating new Connect accounts.
We worked with UX writers to create clear single sign on messaging throughout the flows.
Whatever it is, the way you tell your story online can make all the difference.
USER FLOWS
The product teams developed 21 use cases and user flows to address the scenarios that could occur during after implementation. The main use cases of concern were new and existing customers creating new Connect accounts.
We worked with UX writers to create clear single sign on messaging throughout the flows.
WIREFRAMES
After many brainstorming sessions, and single sign on market analysis we created UI designs for testing the flows with Elavon merchants.
DESIGN SYSTEM AND STYLES
The UX team went to work revising the flows, removing unnecessary steps and working with the application and security to teams to remove steps. Next we A/B tested the new flows against the old flows to determine which ones resonated.
UI DESIGNS
After many brainstorming sessions, and single sign on market analysis we created UI designs for testing the flows with Elavon merchants.
TEST PLAN
I performed a series of moderated usability studies to see how merchants performed in the flows.
I created a research plan, test plan, and test cases. I tested the main use cases with clickable prototypes. I wanted to understand if the account update and linking processes made sense, or if merchants were confused by the flows. Also how they would login to single sign on after account creation.
SURVEY
I performed a series of moderated usability studies to see how merchants performed in the flows.
After identifying th top eleven errors that occured in the system, I created and sent out a survey to over 100 Merchant Planners to gauge their understanding of the error messages. I received over 80 responses which gave us a lot of data to work with, but I still needed to understand the “why” behind the problem. I decided to conduct a design simulation with users.
DESIGN SIMULATION
I performed a series of moderated usability studies to see how merchants performed in the flows.
I asked 9 Merchant Planners of difference experience levels to meet with me for the design simulaton. I also invited my product managers so they could see first hand the errors users were encountering. I walked the Merchant Planners thorugh each error to determine their understanding of the erros and any workarounds they had. I also asked for soluyions that would help them recover.
AFFINITY MAP AND PRIORITIZATION
I performed a series of moderated usability studies to see how merchants performed in the flows.
After the design simulation I met with my team to affinity map the data and to prioritize which errors we wanted to tackle first. I conducted a 2 x 2 pritotization excersize and we decided to first address the errors that presented the highest pain and took the least amount of time to implement.
KEY FINDINGS
83% of users logged in correctly after setting up SSO.
This was important because
Users understood the flows.
Most clinics desire to spend more time in student outreach.
Making clinics more efficient would not have the most impact. Not now, at least.
“College clinics don’t have an efficiency problem, but a capacity problem.”
“Students often will come in because they are stressed about school, and while their problems are important, they don’t have a diagnosable illness. In reality, better study habits may help them even more than a therapist would”
NEW CHALLENGES
Our legal team asked us to remove the “Connect” branding. What?
Another challenge—one of the biggest yet—is that our legal team told us we could not brand single sign on... at all. The best that we could do was call the process “Single Sign On”. There were many discussions following this decision, and we knew that removing the branding would confuse our merchants and affect SSO account conversions.
Most of our merchants are from an older demographic, and do not know what single sign on is. We tried hard to have the decision reversed, and asked to test the unbranded UI with a pilot group.
OUTCOMES
Support calls increased.
After updating the designs and removing the branding we tested my hypothesis in production, releasing single sign on to a small pilot group. Almost immediately support volume increased as users were confused about the process. We had taken the users out of their platform experience with zero branding—and many didn’t know what single sign on was. Drop off rates were high.
These learnings were vital as it gave UX a chance to make a case for maintaining application branding through the account update process. We also worked with our risk team to simplify the flows—removing unnecessary steps.
NEW HYPOTHESIS
Consistent application branding and even clearer messaging will reduce friction in the flows.
We were tasked with re-working and updating the flows, and while we could not use “Connect” branding we made the case for keeping the user in their platform and product branding they were used to.
We also pushed for pre communication emails to the merchants to let them know about the single sign on account update.
PROBLEM STATEMENT
How might we deliver improve the single sign on experience for our merchants?
I revised my hypothesis and theorized that adding branding back to the experience would improve account update success rates.
PRODUCT ROADMAP
How might we deliver improve the single sign on experience for our merchants?
I revised my hypothesis and theorized that adding branding back to the experience would improve account update success rates.
UPDATED USER FLOW
The UX team went to work revising the flows, removing unnecessary steps and working with the application and security to teams to remove steps. Next we A/B tested the new flows against the old flows to determine which ones resonated.
DESIGN SYSTEM AND STYLES
The UX team went to work revising the flows, removing unnecessary steps and working with the application and security to teams to remove steps. Next we A/B tested the new flows against the old flows to determine which ones resonated.
IDEATION
Our legal team asked us to remove the “Connect” branding.
I conducted a white-boarding session with my product team to ideate on possible solutions. Based on user feedback from the design simulation , our solutions were pretty straight-forward. I decided that quick concept tests would help us validate or invalidate our hypothesis.
USABILITY STUDIES
I performed a series of moderated usability studies to see how merchants performed in the flows.
I created a research plan, test plan, and test cases. I tested the main use cases with clickable prototypes. I wanted to understand if the account update and linking processes made sense, or if merchants were confused by the flows. Also how they would login to single sign on after account creation.
A/B TESTING RESULTS
The new flows were clearer with less steps and friction.
Students have trouble identifying stress factors that can spiral into illness.
Most clinics desire to spend more time in student outreach.
Making clinics more efficient would not have the most impact. Not now, at least.
“College clinics don’t have an efficiency problem, but a capacity problem.”
“Students often will come in because they are stressed about school, and while their problems are important, they don’t have a diagnosable illness. In reality, better study habits may help them even more than a therapist would”
“The solution may not actually be an app. Whatever it is, it must be where the students live”
“Students need a primary prevention solution - something that catches the problem before it arises.”
PROTOTYPE
Work has started on a true platform connectivity experience.
Development work has started on the improved experience, and will be released in August 2021. I have also started working on the connection points between the applications to help create the One Elavon vision in the UI..
OUTCOMES AND LEARNINGS
Work has started on a true platform connectivity experience.
Support calls increased.
I next created a concept testing script and presented the revised errors to a new set of Merchant Pkanners, recording the sessions for caputring data. i affinity mapped the data and our hypothesis was validated. I presented the findings to my product team and mockups to the dev team for implementation.