Skip to main content
adidas — CRM Strategy · Systems · Research

Designing a CRM people don't want to mute

Year 2025–2026
Role Sole designer
Scope Strategy · Research · Templates
Delivered to Senior leadership
536M Emails delivered in 2025 Spread across 33 programmes, each built separately and tracking its own metrics
~20% Of programmes driving most outcomes Engagement, loyalty, conversion — concentrated in a fraction of the programmes. The rest were adding volume without adding value
78% Sent to the lowest-performing phases Of all emails went to the user journey phases that performed least. High effort. Low return. At scale
01 — The problem

Designed for the brand,
not the person

High intent validation with room to grow
% total net sales

2025 Q1–Q3 — Email volume vs Net Sales by journey phase

"You send me too many messages that say almost the same thing."

adidas user, 2025

"You only show up when I'm ready to buy, not when I'm still deciding."

adidas user, 2025

adidas had 33 email programmes running at the same time. Each owned by a different team, optimised for a different metric, and none of them connected to each other. A customer who had just made a purchase could receive a re-engagement email the same day. Someone actively comparing products might get nothing at all.

The data confirmed what users were already feeling: most of the volume was going to the moments that drove the least value. The problem wasn't too many emails. It was that the emails were designed around programmes, not people.

02 — The reframe

From programmes
to moments.

01
Discover

"Help me find what fits me and why I should sign up."

02
Compare

"Help me check if this is the right product, in my size, at a fair price."

03
Decide

"Help me complete smoothly, with the best price, no confusion."

04
Use & Enjoy

"Deliver what you promised, support me, and recognise my loyalty."

05
Re-engage

"If I drift, remind me of meaningful value — not spam."

The shift was conceptual before it was operational. Instead of "which programme should we run?", the question became: where is this person right now — and what do they need?

33 programmes were reorganised into 5 journey phases, each with a clear user intent and a defined communication purpose. The programmes still existed — but now they lived inside a framework with a shared logic for the first time.

33 programmes. 5 phases. 3 templates. Each template encodes the logic, tone, and structure of its phase — so global markets can execute without reinventing the system every time.

Templates by user journey phase — Discover/Re-engage, Validate/Analyse, Decide/Convert

Templates by phase — Discover/Re-engage, Decide/Convert, and Validate/Analyse

03 — The templates

Structure that scales

These are two of the three templates developed as part of the system — shown applied to specific programmes as proof of concept. Each template defines the structure, components, and tone for its phase. Any market team can adapt it without starting from scratch.

Template 01 — Validate / Analyse

Under-served discovery intent

Validate / Analyse email — Abandoned browse
From

Only showing interest once the user is ready to buy

↓
To

Help users be confident they're choosing the right product

Validate / Analyse template — components overview
Component 01

Help on size and fit — surfaces the user's saved size and relevant fit guidance. Removes the main blocker at the comparison stage.

Component 02

Social proof — ratings and reviews positioned at the decision moment, not buried at the bottom. Reduces uncertainty without adding pressure.

Component 03

Personalised content based on interest — product image, price, and availability matched to what the user actually browsed. No generic recommendations.

Template 02 — Discover & Re-engage

Untapped loyalty value

Discover / Re-engage email — Rate & Review
From

Sending messages that feel generic and repetitive

↓
To

Show recognition every time they deserved it

Discover / Re-engage template — components overview
Component 01

Reward layer visibility — points balance, tier status, and available perks surfaced prominently. Users need to see the value they have before they can act on it.

Component 02

Hint to next step — a clear, single action that moves the user forward without overwhelming them with options. Progress framing over pressure framing.

Component 03

Positive reinforcement tone — celebrates what the user has already done rather than warning what they'll lose. Language that builds the relationship instead of leveraging fear.

04 — What changed

The reframe in practice

Before After
Structure 33 programmes, each owned in isolation, each optimised for its own metric 5 journey phases — every programme lives within a shared system
Question "Which programme should we run?" "Where is this person — and what do they need right now?"
Tone Loss-framing, pressure, urgency without value Positive reinforcement — celebrate progress, earn re-entry, never threaten
Templates Each team building from scratch, no shared language Phase templates global markets can localise — the system becomes the briefing
Measure Volume and open rates Relevance and engagement quality — CRM as a trust channel, not a volume channel
My contribution

Sole designer. I owned the full journey: research, system architecture, stakeholder alignment and the cross-functional narrative to get leadership to act. Delivered to senior leadership as the strategic foundation for the 2026 CRM roadmap.

"The real problem wasn't the programmes — it was that nobody was asking who the emails were actually for."

Reorganising 33 initiatives into 5 journey phases was a political act as much as a design one. The volume/value misalignment had been visible in the data for months — what changed was framing it as a design problem, not a reporting problem. Frameworks don't adopt themselves: I spent as much time building the narrative for leadership as building the system itself.

← All work Selected projects Next case → When one design has
to feel local everywhere