Major Appliances
Redesigning how customers discover, evaluate, and confidently buy major appliances on homedepot.com.
This is active work — nothing has shipped yet, but it’s one of the highest-revenue, highest-visibility initiatives at The Home Depot. Internal and external eyes are on it. My focus has been on building a research foundation strong enough to make fast, confident decisions when the time comes to execute — and on advocating for a customer experience that matches the complexity and stakes of buying a $1,200 appliance.
Project Brief
Role: Lead Designer — in collaboration with PM, engineering, UXR, and data science
Platform: homedepot.com — Mobile Web and Desktop
Status: Active — in research, ideation, and early testing phases. Nothing has shipped.
Major Appliances is a $3.71B business — 103M visits, 3.3M orders, an AOV of $1,186 — with a 39.7% online penetration goal and a clear mandate to grow it. The online MA team’s mission: unlock a custom and intuitive shopping experience that offers immersive product discovery, drives clear pricing and event experiences, and supports the fastest and least expensive fulfillment channels.
The problem is that buying a major appliance online is hard, and not in ways that are immediately obvious. It’s not just that the product is expensive — it’s that the research process is long (8–12 days on average for refrigerators and cooking appliances), the stakes of a wrong fit are high, the pricing is rarely transparent, and the experience on site doesn’t do enough to close the gap between “I’m interested” and “I’m confident enough to buy.” The three strategic pillars driving the work: make it easy to discover and buy appliances that fit your needs, speed to customer, and deliver clear pricing and event experiences.
The Challenge
The major appliance purchase journey has a confidence problem. Customers need to know a product will fit their space before they buy — and right now, the tools to help them figure that out arrive too late, are too easy to miss, and don’t follow them as they shop. A customer can spend an hour on the PLP finding the perfect refrigerator, only to realize at checkout that it’s two inches too wide. That “heartbreak” moment doesn’t just cost a sale — it drives returns, which is both a financial and operational problem for the business.
A key finding from the research: approximately 75% of major appliance shoppers are in some degree of “duress” — their old appliance broke, their kitchen is under renovation, their timeline is compressed. These aren’t casual browsers. They have a 24–48 hour purchase window, a high cognitive load, and zero patience for UI friction. Page reloads after filter selections, marketing content that doesn’t answer “will it fit?”, and buried delivery date information aren’t minor inconveniences for this customer — they’re the reason they leave.
Layered on top of the fit problem is a pricing transparency gap. Customers frequently don’t realize that buying a major appliance also means buying installation, delivery, and haul-away — costs that surface late in the funnel and create sticker shock. Competitors who surface all-in pricing earlier convert better with the same customer. The Slalom 2023 research put THD’s “Price Paid” score at 2.6 out of 5 — a score that all-in pricing transparency alone could move to 4.5.
And then there’s the Fit Check specific problem: an early version of the Fridge Fit Check feature was rolled back after a calculation error — a 1″ door swing standard was being used instead of the correct 4.5″ — which drove a measurable increase in fridge returns (~$460K impact). The fix required more than a corrected number. It required a redesigned experience that customers would actually trust.
Approach
The work has unfolded across several parallel tracks — which reflects both the complexity of the problem space and the way the Major Appliances IX team operates.
Research foundation. Two major research efforts shaped the direction: the Slalom 2023 MA Research Report, which established the baseline competitive landscape, customer expectations, and THD’s scoring gaps; and the February 2026 Literature Review, which focused specifically on the “Confidence Gap in Fit” — the pattern of customers abandoning purchases because they couldn’t verify a product would fit their space. A Quantum Metric dashboard analysis (February 2026) added behavioral data: comparing ChatGPT-referred traffic to non-ChatGPT traffic to PIP pages revealed that customers arriving with guided, AI-assisted research context converted at nearly four times the rate (2.84% vs. 0.72%) and generated significantly higher revenue per visitor ($8.24 vs. $4.65). That data directly validated the hypothesis that reducing information uncertainty drives conversion — and quantified the upside of doing it well.
Fit Check work. After the rollback of the original Fridge Fit Check, UX completed a refreshed design addressing both the calculation error and the experience gaps that contributed to it. A two-week A/B test is planned before any return to production. In parallel, the broader “Fit Check: Options That Fit” capability — a highest-priority Q2 ’26 initiative with ~$1.3M projected value — is in active Develop & Deliver phase. The customer problem: when Fit Check tells a customer a product doesn’t fit, the current experience dead-ends. The new capability surfaces alternative appliances that do fit, keeping the customer in the purchase funnel rather than sending them to a competitor.
Discovery and testing. Two design concepts are currently in unmoderated testing: “My Dimensions” (a browse PLP with dimensions pinned at the top, fit indicators on products, and fit tools surfaced at the top of the page rather than buried at checkout), and “Simplified PIP for Duress Shoppers” (an “Emergency Mode” toggle that auto-filters for in-stock/fast-delivery options and shifts the PIP to a Specs-First view with an “Include Required Installation Parts” one-click checkbox). Post-test analysis will determine which direction to pursue. Further testing is planned once closer to final designs.
Workshop findings. A cross-functional workshop mapped the highest-priority improvement areas: Visible Fit Tools (fit compatibility tools that appear early in the shopping journey, not only at checkout), Educational Content (clear “what/how to measure” resources in multiple formats including short video), Technical Performance and Inventory Integrity (fixing filter bounce behavior and ensuring sort/filter functions on accurate, live inventory data), and Persistent User Data (letting a customer’s dimensions follow them across sessions and devices so they never have to re-enter them).
High-velocity shopper research plan. A draft research plan specifically for “duress shoppers” is in development — designed around the insight that this customer’s journey is linear and aggressive, not circular. They aren’t browsing. They are entering the funnel with a buy intent already established and asking: “Which of these three in-stock fridges fits my hole and can be here Tuesday?” The research plan includes intercept surveys, unmoderated usability tests, 48-hour diary studies, card sorting, and AI tool benchmarking against IKEA and Amazon.
Key Decisions
Move fit tools to the top of the funnel
Fit compatibility tools currently live at or near checkout — which means a customer can invest significant research time only to hit a fit problem at the worst possible moment. The design work is pushing to surface fit earlier: at the category PLP, on the product card, and in persistent form across sessions. The February 2026 Literature Review named this the single highest-priority improvement area, and the “My Dimensions” prototype currently in testing is the direct implementation of that recommendation. Logged-in customers would have their dimensions stored in account and surfaced automatically; logged-out customers would have them persisted via cookie.
Design for the duress shopper, not the casual browser
The research consistently showed two distinct customer modes: the multi-week researcher who compares options carefully, and the duress shopper who needs the fastest path to a confident decision. Designing for the average of those two produces an experience that serves neither well. The “Simplified PIP” concept and “Emergency Mode” toggle are attempts to give the high-urgency customer a differentiated path — one that surfaces delivery dates, fit confirmation, and all-in pricing immediately, and strips away the content that makes sense for browsers but creates noise for someone whose dishwasher broke this morning.
Surface all-in pricing before the funnel bottom
Customers consistently cite all-in pricing — including installation, haul-away, and delivery — as a key factor in their purchase decision. The Slalom research found that surfacing this information clearly and early could move THD’s “Price Paid” competitive score from 2.6 to 4.5. The design work is pushing to make total cost visible on the PLP and PIP, not just at checkout — reducing the friction that comes from customers who add to cart only to abandon when they see the full cost for the first time. The “Include Required Installation Parts” one-click checkbox in the Simplified PIP concept is the most concrete implementation of this principle.
Where Things Stand
Nothing has shipped. That’s not a hedge — it’s the honest state of a project that’s moving carefully through a high-stakes, high-visibility space. The Fit Check rollback is a useful data point here: moving fast without the right research and the right calculation is more expensive than moving carefully with both.
What does exist is a strong research foundation, two concepts in active unmoderated testing, a prioritized improvement roadmap that has cross-functional alignment, and a team that understands the customer well enough to know what to build and why. The 2026 strategy is clear: reduce returns due to fit and fuel type, unlock Appliance Match Maker, make it easy to discover and buy appliances that fit your needs. The design work is in service of all three.
The metrics that will matter when this work ships: product conversion (currently 3.23%, with a goal of 3.47%), online penetration (39.7%, targeting growth), and return rate reduction — particularly for categories like refrigeration, where fit-related returns are both measurable and directly tied to design decisions. The ChatGPT traffic data already suggests the ceiling: customers who arrive with better information convert at four times the rate. That’s the benchmark the experience is being designed to match.
Next Project
Explore another project: Pro Projects — an in-platform messaging and collaboration system for Home Depot Pro.