Work/ Case Study
Quick Order
Client: Zoro Tools
Role: Experience Design Lead, Growth
Timeframe: Approximately 2 months, discovery to launch, 2025
Team: 1 PM, 1 engineer, 1 designer (me)
TL;DR:
Zoro had sunset its Quick Order feature in 2022 due to almost no usage, about 2.6 add-to-cart events per month. Competitors including Grainger, MSC, McMaster-Carr, and Amazon Business had bulk order tools that Zoro’s high-frequency B2B buyers were leaving the site to use. I led the relaunch end to end: two studies with 15 participants, a phased crawl-walk-run plan, and AI-assisted prototyping in V0 and ChatGPT alongside Figma. The feature shipped in roughly two months and drove approximately $1.3M in session-attributed revenue within 90 days, measured in FullStory.
The Background
Quick Order is not a new idea at Zoro. A version of it had been on the site for years. In 2022, the team sunset it. The data was hard to argue with. Over a 5-month window before sunset, the feature averaged about 2.6 add-to-cart events per month, with no clear visibility into completed orders or AOV. The team pulled the link from the footer behind a feature toggle and moved on.
The hypothesis that brought it back in 2025: a properly designed and properly promoted Quick Order would drive repeat orders from Zoro’s most valuable buyer segment, returning B2B customers who already manage SKU lists in spreadsheets and ERP exports. Internal signals supported the hypothesis. Multiple PO and CS leaders had flagged the gap in MBR conversations. Grainger, Zoro’s sister company under W.W. Grainger, had been running a Bulk Order Pad with paste and CSV upload for years. Amazon Business, McMaster-Carr, MSC, Uline, and Home Depot all had equivalents.
The question was not whether to rebuild it. It was whether we could build it right this time.
The Challenge
Three things had to be true for this not to fail again.
First, the feature had to actually fit how target users work. The 2022 version was a 10-row table with no upload, no validation, no entry point above the footer, and no integration with how purchasing agents and resellers actually manage SKU lists.
Second, it had to be discoverable. Past placement in the footer was a contributing reason for the low usage. If we shipped a better feature into the same dead zone, we would get the same outcome.
Third, scope discipline. The team was small: one PM, one engineer, and me, on a two-month window. Anything beyond MVP had to earn its place.
One of the largest challenges that I wanted to accomplish was creating a full interactive prototype that customers could test with. It was important to see them interact with certain features and actually run it like they would in real life.
I started out using ChatGPT to build some scaffolding and then moved everything to V0 to polish the mid-fi prototype. Check it out yourself.
Research
I ran three studies, fielded between May 12 and May 28, 2025.
Study 1: Competitive Analysis. Performed UX analysis on competitor e-commerce companies already performing bulk ordering. Walked through Uline, Grainger, MSC, or Amazon Business and noted how they actually do it. Some kept the flow simple, and others were more enhanced experiences.
Study 2: Competitor UX Study. N:4, For our competitor study, we fielded research in late May 2025 with four participants who make bulk purchases. They evaluated quick order features from Uline, Grainger, and MSC to help us understand their sentiment and procurement workflows.
Study 3: Unmoderated usability test. N=11, split between Zoro and Zoro-like customers (N=5) and non-Zoro customers (N=6). Respondents worked through an interactive prototype I built using AI tooling. Screened for B2B buyers who order in bulk and have used a Quick Order-style feature(s).
A few findings shaped the design directly.
Manual entry won on trust. 6 of 11 respondents said they would enter items by hand even when given upload options. The theme was accuracy. They did not trust that an upload would catch their errors. Spreadsheet upload still had a place, but for a smaller, more confident subset.
Speed was real and measurable. Respondents took less than 2 minutes on average to add 5 items and confirm. That set a usability benchmark.
More data, not less. 9 of 11 wanted more visible product detail per row. Image, brand, MFR number, stock status. The instinct to keep the row minimal was wrong.
The label was a coin flip. “Quick Order” and “Bulk Order” tied at 4 of 11 each, with “Order by item” pulling 3. We went with “Quick Order” for speed-of-comprehension reasons and because it matched the resurrected internal name.
PO and email are still primary. 10 of 11 still use POs, 8 of 11 still use email, and some send images. Quick Order was never going to replace those workflows. It had to live alongside them.
The desirability test came back with “Efficient, Effective, Professional” as the top respondent-chosen words, which gave Marketing a vocabulary to launch into and product to anchor to in leadership conversations.
Strategy
I built the team an Experience Canvas covering hypothesis, problem statement, key path scenarios, ideas, MVP scope, metrics, stakeholders, and decisions. The Canvas became the single document where tradeoffs got made.
The biggest tradeoff was a phased plan. Rather than one launch, I sequenced two: a Crawl Plan MVP and a Walk Plan.
Crawl Plan, what shipped:
- Standalone Quick Order page, accessible from the main header (no longer buried in the footer)
- 5-row entry form, expandable
- Manual entry with inline validation (out of stock, backorder, item not available, limited stock, exceeds available qty)
- CSV upload via downloadable template
- Live subtotal and Add All to Cart
- Analytics event schema, so we could actually measure this time
Walk Plan, deferred:
- Typeahead and predictive search by Zoro number, MFR number, UPC, or keyword
- Top-nav short-form entry
- Recommendation carousel based on entered items
- Cross-promo into Lists, Customer Hub, and My Account
- Image and quick view per row
- List import and PO drag-drop
What I cut from MVP and why: typeahead got pushed because current Zoro typeahead does not handle Zoro G-numbers well and would have required search team backlog work I could not parallelize. Top-nav short-form was visually attractive but added a second entry surface to test before we had data on the primary one. Recommendation carousel was a Phase 2 feature contingent on adoption data.
Giving the customer what they need
The final MVP design hits the patterns competitor users already understood, then adds Zoro-specific affordances.
Key design decisions:
Header-level entry point.
“Quick Order” sits next to Support in the utility nav. No more footer-only access.
Two parallel input modes on one page.
Manual row entry on the left, CSV template download and upload on the right. The research said manual would dominate. The design did not punish people who preferred upload.
Row-level validation messaging in plain language
“Item is not available at Zoro. Please try a different product number.” “Your item exceeds the current stock. Some of your items may be backordered.” “Limited Stock. Reduce qty to add item to cart.” Each error tied to the specific row, not a banner at the top.
Persistent subtotal and item count
Added to the header strip so users could track the value as they built the order.
Spreadsheet success and partial-success states
Successful upload, upload with some errors, full failure. Each had its own banner pattern with row-level error indicators preserved.
Mobile-responsive layout shipped with MVP
Dedicated mobile flow deferred pending usage signal.
I prototyped in Figma, V0, and ChatGPT, and used the V0 interactive prototype for the unmoderated usability study. That AI-assisted workflow is part of why we hit a two-month timeline with a three-person team.
Results
The feature is live at zoro.com/quick-order.
Within 90 days of launch, sessions where Quick Order was used generated approximately $1.3M in attributed revenue, measured in FullStory. This represents whole-session revenue from buyers who engaged with the feature, not incremental lift versus a holdout.
The qualitative validation also held up. Desirability testing surfaced “Efficient, Effective, Professional” as the top respondent-chosen descriptors, language Marketing now uses in cross-promotion.
Post-launch discovery surfaced clear Phase 2 bets the team will scope:
- Typeahead with multi-field search across Zoro #, MFR #, UPC, and keyword
- Quick view per row with product image
- Save entries and prevent navigation data loss
- Switch and Save pack and multiple substitution
- Import from existing Lists and re-quote functionality
- Mobile scan-to-order, Amazon Lens-style
What I owned vs. what the team owned
I owned discovery, both research studies (recruit, script, moderate, synthesize, report), the Experience Canvas, the crawl-walk strategy, all MVP designs and prototypes, usability testing, and the Phase 2 vision. Product Owner owned product strategy, prioritization decisions, and stakeholder alignment. Our engineer owned implementation. XRD Manager and Director, Product Management, provided executive sponsorship.
Key Aspects:
- 0-to-1 relaunch of a sunset feature
- Mixed-method research, N=15 across two studies
- AI-assisted prototyping in V0 and ChatGPT
- Phased crawl-walk-run delivery
- WCAG-conscious validation patterns
- 2-month total cycle, 3-person team
Technologies Used:
- Figma
- V0
- ChatGPT
- Miro (Experience Canvas)
- Usertesting.com
Are you ready to move the world together?
If you are looking to expand your team or even need support on a single project lets schedule time to chat.
To comply with my non-disclosure agreement, I have omitted and made unclear confidential information in this case study. All information is my own and does not necessarily reflect the views of Zoro Tools or other affiliates.