Explorations and proposals

Password Reset and More Accessible Account Recovery

Good Delivered

The tool enabled associates to reset passwords using basic identifying information, including associate ID, date of birth, and store location. Interviews revealed that many associates did not have their ID immediately available; some preferred speaking with a person for help.

Better Proposed and Best Explored

We considered stronger verification through an additional independent factor, then explored recovery options that could be available when associates needed them: rotating codes, physical identification, fingerprints, facial recognition, and voice recognition.

The research exposed dependencies beyond the interface. Text or email codes required standardized contact infrastructure that did not exist at the time. Physical and biometric approaches raised privacy and device-management concerns. Background noise and microphone availability or quality made voice recognition impractical in many locations.

Decision and Status

These options remained ideas informed by associate interviews and initial discussions. We did not build or test a prototype. As the project ended, we prioritized delivering the basic reset tool. The opportunity to make recovery more accessible and strengthen verification remained unvalidated.

A Familiar Transition for Invoice Matching Users

Better Delivered

The modernized Transfer Invoice Matching System retained the informational inputs, tab sequence, field order, and text-based workflow of the green-screen system. Many users had five, ten, or twenty years of experience. Based on my recollection, onboarding could take approximately six months to a year before someone could work independently.

Best Explored

We tested an optional dark view that recreated the familiar background, green text, and typography through styling changes while preserving the modern application’s functionality. Associates could turn the view on or off. It was intended as a temporary transition aid.

Evidence and Decision

Approximately ten participants across in-store and supply-chain roles tested invoice creation, editing, and deletion. Participants included in-store relationship managers working with large clients who feared that modernization could diminish the value of their accumulated expertise.

Creation was the strongest task. Editing was also generally successful, although some information was difficult to find; deletion was substantially easier. These observations describe the tested experience and do not establish a measured advantage of the optional view.

Participants appreciated the familiar view, but testing did not show a substantial improvement in task success or functional outcomes sufficient to proceed. The view was not delivered. We retained the established keyboard workflow in the modern application.

The lesson was to preserve the interaction patterns users relied on while evaluating visual familiarity on its own merits. Positive preference alone did not establish enough additional benefit.

Building an Internal Shipment Planning Capability

Better Delivered

The product relied on an external shipment-planning engine with a recurring licensing cost. It supported large-scale vendor shipments to supply-chain hubs.

Best Proposed

I proposed reserving one development story every two-week iteration to build the foundations of an internal engine while continuing to use the existing service. The aspirational target was roughly 80–90% of the external engine’s container and truck utilization within one to three years. This was a proposed target, not a tested result or validated delivery estimate.

My hypothesis was that planning tailored to Home Depot’s large-scale products could reduce wasted space and transportation costs, shorten delivery timelines to supply-chain hubs, and ultimately differentiate us from competitors using the same external engine. That assessment reflected my project experience rather than a publicly verified limitation of the vendor product.

Decision and Status

I discussed the proposal with Engineering and Product partners. Product declined because allocating one story each iteration would delay key features. The engine remained unbuilt, and its expected savings and competitive advantage were not validated.

Potential financial value would need to account for development, maintenance, infrastructure, and transition costs. An initial experiment could compare both engines on the same shipment set, measuring capacity utilization, transportation cost, and delivery time. This is a suggested validation approach, not work performed.

Extending Credit Beyond the Application

Work Completed

Before leaving for the Appalachian Trail, I independently extended the design-sprint demonstration over the following two weeks. I completed the remaining workflows for all credit types, including applications, account lookups, and application-status research to determine whether an application had gone through.

I adapted the demonstration across specialty-area desktops, self-checkout, and associate-assisted checkout registers. Shared components and interaction patterns gave the three in-store platforms a consistent experience. This extension was self-initiated and completed without additional team support; it was a demonstration, not a production rollout.

Exploring Best

I developed a ten-point proposal to extend the application direction. The recommendations below are a selected subset; the original full list has not yet been recovered.

A Faster Application with Customer Control

I proposed a target of approximately 30 seconds to apply. Customers would enter a small set of identifying details, such as name, date of birth, and Social Security number, and give explicit permission to retrieve available information. Eligible fields would be prefilled for customers to review, correct, and confirm before submission.

The intent was to remove repetitive entry without obscuring how customer information was used. The 30-second target was aspirational. The availability and suitability of bureau information for prefilling, identity verification, security, consent, and financial-partner requirements remained to be validated. No five-minute baseline or measured time saving is claimed.

Consistent and Accessible Experiences

I proposed a Payments extension of the enterprise design system, with shared branding, colors, components, and interaction patterns. The direction included adjustable text size, high-visibility presentation, screen-reader support, and appropriate audio, with resilient layouts for registers, touchscreens, associate phones, and customer phones. These are recommendations; the demonstrated three-platform coverage does not establish accessibility conformance or mobile implementation.

A Relationship Beyond First Use

The proposal explored mobile-wallet access, ways credit could support an ongoing customer relationship and loyalty, and an open-loop credit option usable outside Home Depot. These concepts required financial and technology partner alignment; customer benefits, feasibility, and business outcomes were not established.

Handoff and Status

I presented the proposals to the team and leadership through Manager and Senior Manager levels, discussed them informally, and handed off a presentation explaining all ten points. The aim was to strengthen the application’s position against an internal alternative from another department. I do not know whether any recommendations were adopted or explored after my departure.

Attribution: Patrick Schumacher’s firsthand account. Completion timing is approximate; future capabilities and expected benefits remained proposals.