The Store Development Walk (SDW) is a quarterly audit used by The Home Depot’s regional and district leadership teams to assess individual store performance and leadership growth. Historically, this process relied on cascading progress reports—from store leaders to district managers, then up to regional leaders—each requiring manual review and interpretation.
However, this volume of information made it difficult for regional leaders to identify trends and struggling stores quickly. The challenge was to introduce AI-generated summaries and insights into the MyView platform, streamlining this reporting process while still respecting the context and intent of the original human-written game plans.
Our team inherited the SDW Walks initiative midway through its lifecycle. The objective was clear: build a scalable, AI-driven dashboard experience within MyView that aligned store, district, and regional perspectives. I led UX research efforts to understand the specific data needs, tone expectations, and dashboard architecture preferred by each persona.
Recognizing the fragmented starting point, I paused high-fidelity work and refocused our team on low-fidelity wireframes and persona-based journey mapping. We honed in on minimal viable functionality, emphasizing language, data hierarchy, and interaction design for our primary users—regional leaders.
We launched with too many user types in scope—regional, district, and store leaders—alongside multiple business stakeholders and an evolving AI toolset. This wide scope made research efforts feel fragmented and usability testing hard to scale.
To address this, I:
Conducted short, focused interviews across all personas to prioritize their most critical needs.
Realigned the UX roadmap around regional leaders as the primary user for the MVP.
Worked with product leadership to trim down scope and focus initial prototypes on summary tone, key data points, and high-level visibility needs.
The project had stagnated under another team, and we were handed partial designs, a backlog of research, and the original deadline. Starting fresh would have been easier—but we needed to respect the prior work while rebuilding momentum fast.
To address this, I:
Audited and organized all previous research to identify actionable insights.
Stripped existing designs down to functional components and rebuilt prototypes from first principles, using a new component system.
Restarted co-working sessions with engineers and business stakeholders to reset expectations and tighten the design/dev feedback loop.
We structured testing and iterations by persona, but always with regional leaders in the lead seat:
Regional Leaders: Our initial prototypes focused on these users. We tested tone, summary views, and callouts for stores needing extra attention. We also validated what constituted “actionable AI feedback” from their perspective.
District Managers: We introduced district-level overviews and AI-generated evaluations of game plans. These views allowed DMs to identify which stores lacked alignment or had underdeveloped plans.
Store Leaders: Although not the focus of our MVP, store leaders received visibility into completed SDWs and AI suggestions for improving their own plans. Their feedback helped shape future iterations focused on mentoring and upskilling.
As of early 2025, SDW Walks has been piloted in two regions. Early feedback has been encouraging:
Regional leaders report significantly faster identification of underperforming stores, with AI summaries surfacing critical trends across districts.
District managers appreciate the ability to monitor game plan submissions and quality without manual tracking.
Store leaders value the AI-generated feedback as a support tool—though they request more specific, actionable insights in future updates.
The MVP successfully laid the groundwork for a scalable leadership development framework. Most importantly, it shifted the SDW experience from reactive reporting to proactive, data-informed coaching.
Less is more when dealing with multiple personas. Attempting to serve everyone at once can stall decision-making and dilute impact.
Persona-based wireframes and journey maps outperformed high-fidelity prototypes in early collaboration stages. These tools brought clarity and alignment faster across cross-functional teams.
Prioritizing tone and communication clarity over data density was key. Regional leaders were less interested in advanced data visualizations and more focused on clear, concise insights they could act on quickly.
Inherited work needs reinterpretation, not rejection. By respectfully auditing and reworking previous designs, we retained institutional memory while moving the project forward with fresh energy.
