Back to projects

Different regions had quietly built different navigations.

I co-led the redesign that put them back into one system. The brief was a menu. The real problem was that no two markets structured the same content the same way.

2x

navigation
click rate

1.8x

mobile
engagement

7x

clicks to categories people couldn't find

EMEA & NA Shipped

Expanding to ROW

Back to projects

Different regions had quietly built different navigations.

I co-led the redesign that put them back into one system. The brief was a menu. The real problem was that no two markets structured the same content the same way.

2x

navigation
click rate

1.8x

mobile
engagement

7x

clicks to categories people couldn't find

EMEA & NA Shipped

Expanding to ROW

Back to projects

From
disconnected nav bars
to one global system
that can bend without breaking

I co-led the redesign that put them back into one system. The brief was a menu. The real problem was that no two markets structured the same content the same way.

2x

navigation
click rate

1.8x

mobile
engagement

7x

clicks to categories people couldn't find

EMEA & NA Shipped

Expanding to ROW

TL;DR

The Problem

The global navigation worked mechanically, but each region had hacked it into something different, diluting the brand and hiding key paths to product pages.

The Research

Stakeholder interviews, competitor benchmarking, geo specific card sorting and testing with 100+ users, testing old designs and comparing new ones to find the best fit on mobile and desktop.

The Solution

A mobile first, photo driven nested navigation with a global spine and regional modules so every geo can adapt content without breaking the system.

My Role

Co-led, working closely with engineering throughout. I owned the research programme across five geos, the A/B test design and analysis, the intent model it produced, and the IA rules.

Co-led, working closely with engineering throughout. I owned the research programme across five geos, the A/B test design and analysis, the intent model it produced,
and the IA rules.

Timeline

May '25 - Apr '26

Context

Users said the navigation worked.
The clicks said otherwise.

In interviews people called the navigation fast, intuitive, straightforward.


There didn't seem to be any complaints. Then we pulled the click data. One search bar out-clicked all nine navigation labels combined. Business, services and software each sat under 2%, which is not what a category looks like when people don't want it.

In interviews people called the navigation fast, intuitive, straightforward.


There didn't seem to be any complaints. Then we pulled the click data. One search bar out-clicked all nine navigation labels combined. Business, services and software each sat under 2%, which is not what a category looks like when people don't want it.

The search bar was doing the navigation's job.

The search bar was doing the navigation's job.

Hover to learn more

India Nav

Convoluted, high moto branding, low contrast fighting against already busy content.

North America Nav

Packed with options, similar sounding labels, little visual guidance.

United Kingdom Nav

Heavy navigation, unclear labels like “motoverse” compete for attention.

Brazil Nav

Long list of moto‑branded terms blur together, hard to tell offers & products apart.

India Nav

Convoluted, high moto branding, low contrast fighting a busy background.

North America Nav

Packed with options, similar sounding labels, little visual guidance.

United Kingdom Nav

Heavy navigation, unclear labels like “motoverse” compete for attention.

Brazil Nav

Long list of moto‑branded terms blur together, hard to tell offers & products apart.

India Nav

Convoluted, high moto branding, low contrast fighting a busy background

North America Nav

Packed with options, similar sounding labels, little visual guidance.

United Kingdom Nav

Heavy navigation, unclear labels like “motoverse” compete for attention.

Brazil Nav

Long list of moto‑branded terms blur together, hard to tell offers & products apart.

Every local fix was individually rational. Together they had made the brand incoherent.

What I took from this

When search out-clicks the menu, that's not preference, it's a findability failure with a workaround. I understood that the solution wasn't to design a better menu, it was a system with rules about what stays fixed and what each region gets to move.

What I took from this

When search out-clicks the menu, that's not preference, it's a findability failure with a workaround. I understood that the solution wasn't to design a better menu, it was a system with rules about what stays fixed and what each region gets to move.

Research

It was never a geography problem

Regional preference didn't hold up under any cut of the data, because every market split internally.

When I sorted sessions by what people said before they started clicking, the split lined up with something they had already decided before they ever arrived on the site.

Hover to learn more

Nested
navigation

“I know what I want.”

  • User arrives with a specific device in mind.

  • Wants a fast, direct path without extra steps.

  • Is cautious about curated pages that might hide better options.

Nested navigation
“I know what I want.”

  • User arrives with a specific device in mind.

  • Wants a fast, direct path without extra steps.

  • Is cautious about curated pages that might hide better options.

  • User arrives with a specific device in mind.


  • Is cautious about curated pages that might hide better options.

Family page
based navigation

“Show me what you have.”

  • User arrives with a specific device in mind.

  • Wants a fast, direct path without extra steps.

  • Is cautious about curated pages that might hide better options.

Family page based navigation
“Show me what you have.”

  • Arrives with a need rather than a product.

  • Wants to see the range of devices before committing.

  • Values a layout that makes comparison easy without overwhelming them.

Research

It was never a geography problem

Regional preference didn't hold up under any cut of the data, because every market split internally.

When I sorted sessions by what people said before they started clicking, the split lined up with something they had already decided before they ever arrived on the site.

Hover to learn more

Nested navigation
“I know what I want.”

  • User arrives with a specific device in mind.

  • Wants a fast, direct path without extra steps.

  • Is cautious about curated pages that might hide better options.

  • User arrives with a specific device in mind.


  • Is cautious about curated pages that might hide better options.

Family page based navigation
“Show me what you have.”

  • Arrives with a need rather than a product.

  • Wants to see the range of devices before committing.

  • Values a layout that makes comparison easy without overwhelming them.

What this exposed

The test made it clear that navigation was not about picking a winner, but about designing one system that respects both ways people decide: direct intent and open exploration.

What this exposed

The test made it clear that navigation was not about picking a winner, but about designing one system that respects both ways people decide: direct intent and open exploration.

Design

We tested two designs that were really two different bets on how people buy.

The homepage wasn't broken. It was invisible.

We took the three most promising patterns into A/B tests across five regions, on both mobile and desktop. Nested and family page models were not just aesthetic choices. They represented two different bets on how people decide what to buy.

Path A – Nested list

Path A –
Nested list

This menu opened into suggested products, and directly taking you from navigation to product page without stopping anywhere in between.

Path B – Routed through device franchise

This sends you to a lineup overview first so you could see everything before choosing. Allows user to explore the all devices in franchise.

How did we turn two behaviours into one navigation?

So we created a third solution, Path C. It came out of the testing once we could see that each model was serving a different behaviour rather than a different region.

What the system does

  • Keeps the nested navigation as the main path for people who arrive knowing exactly which device they want.

  • Anchors the nav to recent and hero devices, so decisive users land quickly where conversion actually happens.

  • Adds a clear “See all [family] devices” entry that opens the full family page for people who prefer to explore the lineup first.

How did we turn two behaviours into one navigation?

So we created a third solution, Path C. It came out of the testing once we could see that each model was serving a different behaviour rather than a different region.

What the system does

  • Keeps the nested navigation as the main path for people who arrive knowing exactly which device they want.

  • Anchors the nav to recent and hero devices, so decisive users land quickly where conversion actually happens.

  • Adds a clear “See all [family] devices” entry that opens the full family page for people who prefer to explore the lineup first.

What changed in my approach

The right answer was not a new pattern. It was a spine and a branch. Nested navigation became the backbone for fast decisions, and family pages became an optional layer on top, available when users want to slow down and explore.

What changed in my approach

The right answer was not a new pattern. It was a spine and a branch. Nested navigation became the backbone for fast decisions, and family pages became an optional layer on top, available when users want to slow down and explore.

The Pivot

70% percent of our traffic was on mobile and we were still designing desktop first.

The homepage wasn't broken. It was invisible.

Our process started with a desktop comp and treated mobile as something you resized into at the end, which made less sense the longer I looked at where people actually were.

I pushed to invert the order, and from that point every model got drawn and tested on a phone before anyone opened a desktop file. Of everything I argued for on this project, that one had the largest effect on the final design, and it wasn't a deliverable at all.

Then the roadmap outgrew the design

A new phone family was coming, the first in years, and accessories were moving from a side category into a real part of the story.

Path C held the existing lineup well, but it held it because I had designed against that lineup, and every family had a home for the simple reason that I knew every family.

What this exposed

What the research has
exposed

Our first navigation worked well for the world we had designed with no space for growth. The moment a new family entered the picture, that assumption broke.

How the problem was fixed

We reworked the navigation into a spine with open slots. We defined a pattern that could host new franchises and accessory groups without rewriting the IA each time.

Where the test left us

The first version fit the catalogue we had. The second one had to survive the one we hadn't seen yet.

Then the roadmap outgrew the design

A new phone family was coming, the first in years, and accessories were moving from a side category into a real part of the story.

Path C held the existing lineup well, but it held it because I had designed against that lineup, and every family had a home for the simple reason that I knew every family.

What this exposed

Our first navigation worked well for the world we had designed with no space for growth. The moment a new family entered the picture, that assumption broke.

How the problem was fixed

We reworked the navigation into a spine with open slots. We defined a pattern that could host new franchises and accessory groups without rewriting the IA each time.

Where the test left us

The first version fit the catalogue we had. The second one had to survive the one we hadn't seen yet.

The Solution

What does the navigation look like after we rebuilt it for scale and regional needs?

How does it work?

The primary navigation leads with a nested list of device families and key products, anchored to recent and hero launches. Inside each family, users see specific devices first, with a clear “See all [family] devices” link that opens the full lineup when they want to explore.

The Solution

What does the navigation look like after we rebuilt it for scale and regional needs?

How does it work?

The primary navigation leads with a nested list of device families and key products, anchored to recent and hero launches. Inside each family, users see specific devices first, with a clear “See all [family] devices” link that opens the full lineup when they want to explore.

Why this holds together

Decisive users can jump straight from the nav to a product page. Explorers can step into a family overview without losing their place. New franchises and accessories plug into the same pattern, so the system stays stable even as the catalog grows.

Why this holds together

Decisive users can jump straight from the nav to a product page. Explorers can step into a family overview without losing their place. New franchises and accessories plug into the same pattern, so the system stays stable even as the catalog grows.

Anchored in what is new

Hero and recent devices are pinned in the nav so the freshest launches are always one click away.

Anchored in what is new

Hero and recent devices are pinned in the nav so the freshest launches are always one click away.

Ready for the next launch

New franchises and accessories drop into predefined slots instead of forcing a redesign.

Ready for the next launch

New franchises and accessories drop into predefined slots instead of forcing a redesign.

Consistent structure,
local catalogue

Every geo shares the same navigation pattern, but can surface its own offers and priorities inside it.

Consistent structure,
local catalogue

Every geo shares the same navigation pattern, but can surface its own offers and priorities inside it.

Consistent structure, local flavour

Every geo shares the same navigation pattern, but can surface its own offers and priorities inside it.

Consistent structure, local flavour

Every geo shares the same navigation pattern, but can surface its own offers and priorities inside it.

Impact

Navigation click rate was 2x

Comparing two two-week windows either side of launch, across EMEA and North America.


Mobile interaction moved further than any other surface, which mattered because that was already where most of the traffic was.

0.6% → 4.1%

business, home devices and software support, the three sections almost nobody was reaching

5% → 10%

smartphones,
which is the path that was most important for sales

7x

combined clicks to sections that saw very little interactions before, increasing discoverability

Learnings

I stopped designing for the
catalogue in front of me.

Design for the future, not just for the moment

Path C failed because I built it around the lineup in front of me, and lineups change faster than structures do. I look for the smallest set of rules that survives the next three launches instead.

What people call easy and what people complete are different measurements.

They praised the family pages, never scrolled to the CTA, and failed the task anyway. I stopped treating stated ease as a result.

Search share is a diagnostic, not a feature.

Ours climbed after launch, which either means people prefer search or means they're still failing in the menu. I never separated the two, and I'd want that instrumented from the start next time.

The version that shipped isn't the one I was proudest of, it's the one still standing after three more launches, and I'd make the same trade again.