Skip to main content
adidas — Internal Tooling · Scale · UX

Content at the speed of a global brand

Year 2024
Role Lead Designer
Scope Research · Systems · Product
Teams 4 cross-functional teams
+27.93 User satisfaction score in 6 months From 54.72 to 82.65 — across global and market users
54% Faster content creation Compared to the previous third-party CMS
4 Teams co-designing A flexible global-market model with shared ownership
01 — The challenge

Built to publish,
not to think

adidas publishes content across hundreds of touchpoints at the same time — campaign pages, emails, app banners, partner sites. The problem was that the tools used to do that had piled up over the years. By the time this project started, teams were juggling 11 different systems just to get one campaign out the door.

Most of their day wasn't spent creating. It was spent moving between tools, waiting on handoffs, and adapting the same content manually for each market. The creative work happened in between.

"I spend half my day doing things the tool should do for me."

Content Manager — market shadowing session

"By the time we publish, the campaign window has already shifted."

Market Activation Lead — shadowing session
02 — Research

Complexity was
the work

We started by sitting with the people who used the tools every day. We ran shadowing sessions with market teams and mapped every tool touched in a single campaign lifecycle — from the initial brief to the moment something went live. The pattern was impossible to miss: people had learned to work around the system, not with it.

01
Tools mapping

11 systems involved in a single campaign — planning, creation, localisation, scheduling, publishing. Many overlapping, none connected. Each one a different way of doing the same thing.

02
Market shadowing sessions

On average, teams switched between 6 to 8 different tools to publish one content piece. Adapting for local markets wasn't a structured step — it was something each team figured out on their own.

03
Customer journey mapping

Personalisation always happened after the content was already made — templates duplicated and edited by hand for each audience. There was no way to think about variation while creating.

04
Opportunity trees

We turned all of this into a set of product requirements grounded in real feedback — not assumptions. What users actually needed, what the old system could and couldn't do, and where the real blockers were.

Tools mapping across the full campaign lifecycle

Tools mapping — 11 systems across planning, concept, adaptation, activation and measurement

Market shadowing sessions — pain points, nice to have, possible solutions

Cluster insights from market shadowing sessions — pain points, what works, possible solutions

Customer journey map across content lifecycle

Customer journey map — content lifecycle from early access to publication

Opportunity trees — product requirements grounded in user research

Opportunity trees — requirements grounded in real feedback, not assumptions

1 / 4
03 — The solution

One place for everything

The new CMS brings creation, localisation, and publishing together in one place. You can see what's live, what's waiting for review, and what's blocked — without checking five different tools. Personalising content for different audiences is something you do while creating, not after the fact. And every market team works from the same foundation, adapting what they need to without starting from scratch every time.

Operations Module — final design, personalisation interface with recommendation experiences

Operations Module — recommendation experiences, strategy, country selection and live preview

04 — What changed

The shift in practice

Before After
System 11 tools for briefing, creation, localisation, and publishing — none of them connected A redesigned core platform, better integrated with the existing ecosystem — one place for creation, localisation and publishing
Localisation Each market team rebuilt global assets from zero — inconsistent results and duplicated effort Local teams adapt, not rebuild — they work within a shared structure, so global intent always comes through
Personalisation Handled after content was finished — templates copied and edited by hand for each audience Built in from the start — you set up audience variations while creating, not after
Visibility Campaign status tracked in spreadsheets and long email threads — nobody had the full picture Everyone sees the same thing — what's live, what's in review, what's blocked, all in one view
Speed Content took longer to produce than the campaign window it was made for 54% faster content creation — less time on tooling, more time on the actual work
My contribution

Lead Designer across discovery, systems architecture and delivery. Led cross-functional alignment across global and market teams. Defined the component model and the global-local handoff framework that shipped to production.

"The tools weren't the problem. The mental model of how content gets made was."

Eleven tools had accumulated over years because each team had optimised for its own workflow. Making the case for a better core wasn't about showing the inefficiency — leadership already knew. It was about making the cost of fragmentation feel real, and giving teams a model they actually wanted to work inside. At this level, architecture and persuasion are the same job.

← Previous case Making energy management
feel effortless All work → Selected projects