Actionable Shelf Insights
Store leaders at The Home Depot managed On-Shelf Availability (OSA) across 6–8 disconnected tools. I led the research and design of a connected reporting experience: a quick overview in myView, the app store leaders use for daily tasks and store walks, with drill-downs to the SKU in Store Pulse, the company's detailed store-performance reporting tool. It went from a one-store pilot to about 2,000 stores in about three months.
Project Overview
Role: Senior UX Designer (sole designer)
Led the end-to-end UX process for Store Pulse, including user research, UX strategy, interaction design, prototyping, and usability validation.
Team: PM, data analytics, engineers
Timeline: 6 months
Platforms:
-
MyView: A centralized experience providing store associates with daily operational insights, priorities, and tasks.
-
Store Pulse: A reporting application providing store and field leadership with detailed performance insights to support operational decision-making.
Target Users: Store Managers, Assistant Store Managers, Department Supervisors, District Managers, and Regional Managers.
Research & Design Methods:
-
Stakeholder interviews and user interviews
-
Existing system and workflow walkthroughs
-
User flow mapping and interactive prototyping
-
Moderated usability testing
-
Data-informed design validation
Problem Statement
Keeping shelves stocked is a store-wide effort, from opening to close. Store leaders are accountable for OSA, but they had no dedicated way to spot issues, check execution, or act.
Too many tools
One Operations ASM relied on 6–8 tools to manage OSA and worried that missing one would hurt his store.
The wrong signal
Leaders' OSA overview came from Voice of the Customer scores. As a District Manager pointed out, only about 5% of customers respond, and high scores don't mean full shelves.
Invisible work
Sidekick generates packdown tasks from sales data, but leaders couldn't see what was left untouched without checking each department one by one. "It's like operating blindfolded when it comes to task completion." (Operations ASM)
Points over shelves
Associates earned Sidekick "gems" and some chose easy logged work over directed tasks. As one Department Supervisor said, the system depends on associate integrity, so leaders had to verify the work themselves.
HOW MIGHT WE
give store leaders a high-level view of the OSA metrics that matter most in myView,
and let them act on it?
My Role
I was the only designer on a small team with a PM, data analytics and engineers, and I owned the work from research to rollout.
-
Research: led multiple rounds of interviews and usability testing with Store Managers, Assistant Store Managers, Department Supervisors, and District Managers.
-
Strategy: defined which six OSA drivers shipped first, and which metrics to hold back until the data could support them.
-
Design: designed the Store Pulse OSA drill-downs, and added the link from myView to Store Pulse that connects the two tools.
-
Alignment: shared a readout with business partners after every round.
-
Cross-team: studied Sidekick in depth, designed its SKU/Bay and Purge reporting, and consulted a Sidekick engineer and UX on linking SKU data to classes. That work was carried forward after I moved to the MET tasking team.
Research
Before designing anything, I needed to understand how OSA actually works inside a store.
I ran multiple rounds of interviews and usability sessions with Store Managers, Assistant Store Managers, Department Supervisors, and District Managers. I also worked with data analytics to understand what data existed and what was reliable. I opened the conversation with the Sidekick team, and that work was carried forward after I moved to the MET tasking team.
Mapping the operation. I traced the four stages that keep a shelf full, and the tools used at each one:
Understanding each driver. I then went deep on the tools behind each metric: Sidekick tasks and points, Overhead Imaging, Overhead Management location tags, Urgent Tasks from online-order cancellations, and the on-hand approval process. For each one I documented who uses it, what goal it serves, and where leaders lost visibility.
Key Insights
The company didn't lack reports. It lacked reports that told leaders what to do next.
As one regional operations leader (RDO) put it, Home Depot is great at producing reports but struggles to make them actionable. That reframed my goal from "more data" to "actionable OSA drivers."
91%
of store leaders rank OSA right after safety and sales
75%
say Sidekick gems don’t matter. Full shelves do.
Labor or effort?
Without task data, leaders couldn’t tell which was the problem.
Coach, not just monitor
Leaders wanted to know who needed training.
Broad, then deep
A quick daily view, with detail one tap away.
Strategy: what to build first
I narrowed nine candidate metrics to six for the first rollout, and deliberately held back the ones we couldn't support with reliable data.
I weighed each candidate on four criteria: direct impact on business goals, feasibility and complexity, user needs, and available resources.
Holding back the Health Score was a deliberate trade-off. A single headline number is appealing, but a score we couldn't stand behind would erode trust in every report.
The system: two tools, one path
Instead of introducing another tool, I focused on clarifying the role of each existing tool and creating a more connected experience.
MyView and Store Pulse were developed by separate teams, which created confusion among store leaders. Many perceived them as competing tools rather than complementary solutions.
Department Supervisors relied on MyView but found Store Pulse difficult to learn. Meanwhile, store managers and assistant store managers valued the deeper operational insights Store Pulse provided.
My hypothesis was that store leaders should start their day in MyView to get a quick snapshot of store performance, then move to Store Pulse when deeper insights are needed. To create a seamless experience, I designed each OSA card in MyView to link directly to its corresponding report in Store Pulse.
Each tool does one job, so leaders get a quick read every morning without losing access to the detail.
Design Decisions
Every design choice traced back to a pain point leaders described in research. Here are five that mattered most.
1. Show the work Sidekick hid
Leaders couldn't see untouched tasks, so they couldn't tell a labor gap from an effort gap. I added a count of incomplete tasks to the Sidekick card in MyView, then built full task detail in Store Pulse: the department, the associate, the bay, the time completed, and the SKUs packed down.
2. A neutral progress bar
I chose not to label green as success or red as failure. The bar simply shows completed tasks against tasks sent. Sidekick's task volume swings through the week, with Mondays heaviest, so I wanted the bar to inform rather than judge.
3. Completion % comes last
In the Store Pulse department table, I put Completion % in the last column on purpose. Departments get different task volumes, tasks hold different numbers of SKUs, and some packdowns are harder than others. Leading with the percentage would invite unfair comparisons.
_2x.png)
4. Turn data into a coaching moment
For No Location Tags, leaders wanted to know who needed training. I added a list of associates who created untagged pallets more than twice, with a link to each tag's detail. One Operations ASM noted that whoever created a tag leads you to the team that placed the pallet.
5. Make freshness visible
Old overhead photos are useless for finding stock. In the Overhead Imaging report, I added Image Needed, Last Imaged and Days Old, flagged once a photo is more than 7 days old. Leaders can drill from department to aisle to bay to see exactly where to send someone.
_2x.png)
I also aligned the layout with the store's existing scorecard, splitting Sidekick SKU/Bay from Purge, so leaders could find numbers where they already expected them.
Testing and iteration
Usability testing came back overwhelmingly positive, then surfaced a data problem I had to design around.
Leaders responded strongly to the flow prototypes. They also pushed for more: which day of the week had the most untouched tasks, what happened while they were out sick, and detail down to the SKU and the tag.
The challenge was that some of the data in the prototypes wasn't available yet. Rather than delay the launch, I built a Data Verified version that used only the data we could deliver reliably, and tested it again. Where detail wasn't ready, I was transparent about it:
-
Urgent Tasks showed an overview only, because the feature was moving from Smart List to Sidekick.
-
On-Hand Auto Reject Rate had no drill-down yet, because the legacy Store System couldn't supply the data.
A District Manager summed up the priority: the MyView overview was a good start, but the link to Store Pulse mattered, because stores couldn't wait to investigate.
Store Manager 1
Very easy to understand.
Assistant Store Manager 2
If we can have all this in Store Pulse, I don’t need to use Sidekick.
Store Manager 3
I really appreciate this Associate list.
Assistant Store Manager 4
This will be a huge help for the store.
What store leaders said in the testing
Rollout and impact
The OSA report went from one store to about 2,000, and it's been available to every store leader since.
~2000
stores with the OSA report
4 of 6
first-release drivers still tracked in Store Pulse today
Daily
store leaders tell me they open Store Pulse many times a day
We rolled out in stages so we could learn before scaling:
-
1 store: QA and functionality testing.
-
6 stores: the MVP, with results reviewed with support-center and field leaders.
-
25 stores: a wider pilot.
-
About 2,000 stores: a three-week expansion to every remaining store.
Before the full rollout, we reviewed the work with the Ops Fulfillment VP and with the SDOs, then submitted it for company-wide approval. Store Pulse reporting followed the MyView launch, expanding the six OSA insights into department-to-SKU detail.
The lasting signal. Internal Store Pulse guidance now ties effective use of the tool to gains in Sidekick execution, overhead imaging and no-location tags, the same drivers I prioritized for the first release. Two years on, store leaders still tell me they love Store Pulse and open it multiple times a day, and OSA is part of that daily habit.
One leader also noted a practical win: completion rates had arrived by email twice a week. With the report in Store Pulse, they no longer had to wait for it.
Two years later
The structure I designed in 2024 is still how store leaders work with OSA today.
The page has kept evolving, with new measures like MET Out Scans and Sidekick Task Actionability. The foundation, though, is the one we set: a small set of actionable drivers, each one a tap away from the detail.
Reflection
The biggest lesson: the most useful thing I designed wasn't a screen, it was the relationship between two tools.
-
Design the hand-off, not just the destination. Department Supervisors found Store Pulse hard to learn, yet leaders now use it many times a day. Giving MyView a clear "start here" role and linking each card to its detail helped make the deeper tool approachable.
-
Trust is a design material. Holding back the Health Score and shipping a Data Verified version kept the reports credible. A report leaders don't trust is a report they stop opening.
-
Measure impact from day one. If I did this again, I'd agree on before-and-after metrics with data analytics at kickoff, so the impact story is as strong as the qualitative one.
What's next. The second phase was planned to add Receiving, Freight and a Sidekick leaderboard. Leaders told us associates are competitive and would be motivated by it, so the design question is how to make competition encourage the right work instead of easy points.
