lululemon logo

lululemon

How I ran product design for lululemon's North American online store: sprints as the default loop, a design system as the foundation, and the navigation, product page and community work built on it.

Timeframe

Jul 2018 to Jun 2020

Company

lululemon

Role

Manager, Product Design, North American E-commerce (UX)

Author

Michael Preuss

Updated

2026-10-27

Read time

4 min

Read with...

Pick an agent

Gemini and Your Claw use copy and paste.

  • 01Led the eight-person product design team for lululemon's North American online store, reporting to the EVP of Ecommerce
  • 02Shipped a design system with a code library: component variation down 60%
  • 03Redesigned navigation and the product page on sprint cadence; NA ecommerce revenue grew 30% over the period, with UX one input

I led the eight-person product design team for lululemon's North American online store during a period when annual ecommerce revenue grew 30%, from about $850M to $1.1B. Our work focused on the design system, navigation and product pages that supported that growth.

As Manager of Product Design from July 2018 to June 2020, I reported to the EVP of Ecommerce and set the team's operating cadence.

July 2018

The store I inherited had too much variation. One-off patterns and near-duplicate components had piled up, so every journey was slow to change and the experience drifted between navigation, browse and the product page as each team shipped. Traffic was moving to mobile and to the product page while the classic path from the homepage through category pages declined.

What I decided

Make design sprints the default loop. Every significant piece of work started with a sprint that put Product Design, Product Management, Engineering and frontline store staff in the same room, with Search Marketing when the problem was findability. Sprints forced the tradeoffs early. On the product page, one caught an early concept that posed a risk to conversion and could have required costly rework; on navigation, one produced the four guest need states the whole redesign was built on. (Guest is lululemon's word for the customer.)

Build the foundation once. We audited every component on the platform with Atomic Design as the lens, documented the duplicates and near-variants, and ran four two-week sprints to ship version 01 of a design system: a Sketch library in InVision DSM and a coded library in Storybook, in separate repositories so it ran as a service rather than a monolith. I led it from the product design side, drawing on ten years of front-end work, and feedback from the design team took it to version 02 quickly. Component variation fell 60%. By the end of 2019 the plan was to resource it as its own product team.

Spend the capacity on the journeys that carried revenue. Navigation and taxonomy, so guests could find what they came for; the product details page, where browsing turns into a purchase; and Community Now, the platform that takes the sweatlife beyond the store.

How it ran

Navigation started with the words. Merchants preferred "crops" and "tanks" as on-brand; a Google Images search for either showed what everyone else means by them. Four need states, from "I know what I want and I know how to find it" to "I have no idea what I want", and three pillars, speed, inspiration and clarity, shaped the structure. The first concept took two weeks; user testing sent us back to make guidance primary and inspiration secondary, and the third round landed. The taxonomy and navigation groups worked so closely that leadership merged them, and a new search platform, taxonomy and navigation rolled out from 2019.

The product page went mobile-first. The first viewport stays out of the way so you can see the product and add it to your bag; about half of guests scroll past it looking for more, so the second viewport educates progressively, icons first and accordions for detail, with outfitting, similar products and personalized recommendations below. Our lead product designer iterated each component with weekly reviews, bi-weekly steering committee demos in clickable prototypes, and quarterly presentations to the EVP and CEO. Requirements became epics in JIRA with A/B tests; most came back positive, and the components were prepared for launch ahead of Q4 2019. The success metrics were add to bag, Voice of Guest, performance, conversion and bounce rate.

Community Now was designed on four pillars: store hubs, classes and events you can book, ambassadors, and a content hub, with a release planned for the end of 2019.

The same team carried program management through Black Friday and Cyber Monday, when execution had to be tight under load.

The tradeoff was depth over breadth. We cut one-off design work and put the hours into the system, the sprint cadence and the cross-functional rituals that scaled across teams.

Results

These outcomes come from ecommerce performance reporting, design system governance tracking and Voice of Guest program reviews from this period. This page is the primary published record of the work.

  • Revenue: North American ecommerce revenue grew 30% over the period, from about $850M to $1.1B. The UX work was one input among many.
  • Consistency: component variation down 60% after the design system; Voice of Guest complaints declined as the experience got more consistent.
  • Velocity: sprints as the default loop and a shared foundation made delivery faster; version 02 of the design system followed version 01 quickly.
  • Operating system: design sprints by default, a design system as the foundation, and a steering cadence from weekly reviews to quarterly executive presentations.

Deep dives

In the order the work was built:

FAQ

Questions this page answers

What did Michael Preuss do at lululemon?

I was Manager of Product Design for North American ecommerce from July 2018 to June 2020, reporting to the EVP of Ecommerce. I led the eight-person product design team for lululemon's online store and set the cadence it ran on: design sprints by default, a design system as the shared foundation, and regular reviews with a steering committee and the executive team.

What problem was the work solving?

Too much variation. Journeys and components were treated as one-offs, so delivery was slow and the experience drifted between navigation, browse and the product page. The fix was to build the foundation once and spend the team's capacity on the journeys that carried lululemon's revenue.

What results came out of it?

North American ecommerce revenue grew 30% over the period, from about $850M to $1.1B, with the UX work one input among many. Component variation fell 60% once the design system landed, Voice of Guest complaints declined as the experience got more consistent, and delivery got faster once sprints were lululemon's default loop.

What did the team work on?

A digital design system with a code library, a navigation and taxonomy redesign built on four guest need states, a mobile-first product details page, and the product design for Community Now, lululemon's community platform. The team also carried program management through Black Friday and Cyber Monday.

Which deep dive should I read first?

Start with the design system, because everything else at lululemon was built on it. Then navigation and taxonomy, then the product page, then Community Now.