Labor is one of your largest investments in any business, especially in retail. Our balanced product team, Store Labor Management (SLaM), had the responsibility of maximizing a return on an investment that is over half a billion dollars.
PROBLEM
(Short on time. Click icon for quick summary)
Pre-Covid: Optimize in-store scheduling for 800+ stores and 30,000+ employees to place the right teammate, in the right place, at the right time.
During-Covid: Improve workload planning at a corporate and store level so stores don't feel overwhelmed with tasks, and have the ability to execute tasks while also engaging customers.
Role:
Responsible for the entire user-centered process from research to executing a solution based on research findings and user testing.
Team:
1 PM, 1 UX Designer, 4 Engineers
Disclaimer:
Due to the type of work domain, I’m limited in the amount of work I can display publicly.
{ Additional commentary in red }
PROCESS & APPROACH
Initially, there were only two of us on the team. One Engineer and myself. The rest of the team wasn't hired yet. The only guidance we received was labor was a large investment with labor costs continuing to rise and maximizing that investment was extremely important.
The domain was too broad and needed more clarification, so my strategy was to:
-
Look at labor in the retail industry as a whole. Reviewing trends, statistics, labor impacts and outcomes, etc.
2. Gather as much data as possible around labor at Dick's Sporting Goods (DSG). Compiling hiring and turnover numbers, employee surveys, pay, tenure, etc.
3. Understand how DSG store labor compares to other retailers in the industry.
4. Share and display how DSG labor is similar or different compared to the industry.
{Sharing and displaying provided transparency and enabled conversations with other product teams to understand how their work impacted store labor, and also help me narrow our teams scope. Most people enjoyed the transparency, however there were a few people who didn't like the findings}
5. Make a list of enterprise contacts who may have input into store labor, and have conversations to determine key stakeholders.
6. Plan research.
7. Interview and observe Stakeholders, Store Management, and store teammates to learn work behaviors, labor planning, tasking, performance measurement, and thoughts around pay and tenure.
8. Use research to extract key insights in order to highlight possible solutions.
RESEARCH
Due to my previous experience working in Retail Management for different companies, I had a good understanding of how retail labor operates. However, the retail industry is constantly changing and evolving and I wanted to get the most up to date news and trends.
{These are just a few of the resources I studied}
After gaining more understanding, I wanted to gather findings specifically for labor in DSG stores.
{Sharing and displaying findings}
Findings garnered:
Over 500 data points
24 Categories synthesized
Over 75 Problems related to labor spanning the entire company
At this point in the process, we added more team members; 1 Product Manager and 2 more Engineers.
{To get others up to speed, the research (interviews, recordings, and other documentation was explained and shared}
The data was good, but we needed to focus on a few key things that would be valuable to the user and business. We presented our research to Stakeholders and our Product Leadership Team. The presentation segued into a Goals/Anti-Goals session that helped create alignment, better defined domain, and clearer boundaries for our team to work within.
PRIORITIZATION
Now with a clearer picture of our direction, we conducted further research on one of the 24 synthesized categories. The category involved scheduling, timekeeping, and attendance.
{Research methods used - Interviews, Observational}
{We also analyzed current internal store reporting on schedules, workload planning, and customer satisfaction scores}
Findings garnered:
Over 700 data points
Over 100 problems specific to scheduling, timekeeping, and attendance
We chose the opportunity with the highest value for both users and business.
After synthesis, we analyzed the problems to see if there were any close relationships and prioritized them according to user frustration, frequency, and business impact
PIVOT
March 2020 COVID hits. Stores close and retail as we know it changes.
Suddenly, shopping behaviors change, store labor and scheduling changes, and our team has to pivot.
With restrictions in place due to Covid, shopping shifted solely to online. The increase of Ship-From-Store (SFS), Buy-Online-Pickup-In-Store (BOPIS), Curbside Pickup, and newer additional processes significantly impacted store labor. DSG now needed a way for stores to plan and manage the surge in workload from online business while not overwhelming a reduced and concerned workforce.
MIRO
Slack
MIRO Boards, Slack, and Microsoft Teams became essential tools for our newly remote team to conduct work. Instead of visiting stores in-person, we conducted interviews over Zoom, Microsoft Teams, and the telephone. Emails and other correspondence needed to be clearer and more direct when sharing findings and insights.
Microsoft Teams
FINDING THE REAL PROBLEM
Throughout the enterprise everyone was adjusting to a new normal and new standards in DSG stores. Because of these changes, our team was given a problem to solve, but we weren't sure if that was the correct problem.
Given Problem: Create workload planning specific to each individual store.
{The thought was if the plan was store specific, then stores could better manage labor costs and improve P&L}
What We Found: Store Managers, who are responsible for managing the hours and tasks, didn't have a problem with the current workload plan they received. The main challenge they faced was the unexpected tasks (workload) that would come from corporate.
The research directed us to focus more upstream on where the workload was created. Allowing us to understand why and how unexpected workload was being sent to stores?
{Workload Journey's}
The more we followed the journey of different tasks (workload) stores were required to complete, the more we saw issues in the process. We also saw frustration from users who are responsible for mitigating all the work being sent to stores.
Current planning, mitigation, and forecasting relied on spreadsheets, emails, and meetings without the ability for different departments to see how their tasks contributed or impacted store labor.
This process put strain on the planners who felt they were constantly overloading the stores with too many planned and unplanned tasks.
Workload Planner
The research informed a solution focused on solving frustrations for workload planners who need a better way to forecast and mitigate workload so stores wouldn't be overwhelmed.
{At this time we recognized there was potential to forecast approximately 75 - 80% of store workload, but that depended on a few other factors we were waiting to get answers for}
DESIGN
Users wanted to be more proactive instead of reactive to potential problems. They also complained about the lack of centralized visibility within the planning process where decisions are made.
Below are a few elements that needed to be visible:
-
Workload
-
Risks
-
Estimated vs. Actual Hours
-
Percentage of Workload already assigned
-
Historical Data
-
Estimated labor costs
Pencil & Paper
Mid-Fidelity
After a few iterations based on feedback and testing, the design solution involved taking current workload and planning elements and visibly enabling the user to see current workload statuses, future workload, and potential risks. Having this visibility allowed for more proactive decisions.
{The solution was a desktop application. Users completed work on laptops and desktops}
High-Fidelity
RESULTS
Even though our users were pleased with the solution's direction, we could only forecast approximately 60% of workload due to factors and processes outside of our control.
Technology was helpful, but this was more of a process and procedure problem that required complete cooperation from several departments at an enterprise level.
We met with our users and stakeholders to discuss the feasibility and potential of the application. We discussed the challenges and trade-offs if processes and procedures were not open for change.
What we accomplished was lauded, but the decision was to pause the project until changes at the enterprise workload planning level could be resolved.
Previous:
Next: