Apparicio

Head of product design

Apparicio

Apparicio

2022

2025

Head of Product Design @ EE / BT Group

Three design visions, a rebuilt profession, and the design language EE runs on today.

A rebrand hands over colours, a logo and a typeface. Someone still has to decide how they behave on a screen.

My role

Head of Product Design

Timeline

2022 → 2025

Team Composition

A profession that grew to 139 designers, 16 reporting to me directly

LINK

ee.co.uk

EE is BT Group's consumer brand, selling mobile, broadband and TV to millions of households through an app, a website and a retail estate. Design sat spread across tribes and squads, with product design, content design, research, service design and accessibility all reporting in different directions.

If you only read one part

EE was shipping steadily, and design arrived at the end of decisions rather than the start. It could improve a screen and never change a plan.

What I led was a change in what design was for, and a direction set far enough ahead to pull the roadmap rather than follow it.

The spacing model, the motion rules and the way the brand behaves on screen are still the definitions EE designs by today.

The problem

The team wasn't short on talent, or headcount, or budget. It was short of a reason.

No one was asking what the product should become because nothing in how design was organised made it anyone's job to ask. A vision that only contains things already agreed is a plan, and plans were not in short supply. So I wrote one, and deliberately pitched it further out than the roadmap reached.

96,000 components hardcoded in production, with no central governance.

A 60% UI bug rate in the app.

The best designers sat isolated inside their own squads.

No metrics for design, so it read as an aesthetic layer rather than a business driver.

No agreed difference between a UX designer, a UI designer and a product designer.

Engineers believed hardcoding was faster than the design system, and on the evidence they were right.

Customers met a different design language in the app, on the website and in the portals.

Projects followed a funding model rather than an experience model.

22% engagement in the design community, unclear progression, ad-hoc training.

Every one of those is a reason not to look up. Fixing them was necessary, but on its own it would only have produced a tidier version of the same team.

Where I started

We could not fix all of it at once without breaking the product, so the work started as questions rather than answers.

01.

How do we apply human-based decisions?

02.

How do we bring users into the decision process?

03.

How do we improve cross-collaboration?

04.

How do we increase design system adoption?

05.

How do we change the leadership layer?

What I did for the vision

Wrote the first design vision in my first six months.

Before restructuring anything, the team needed something to aim at. A design organisation with no picture of the future will optimise the present forever, and that is what was happening.

So I wrote one, and deliberately pitched it past what the company thought it was allowed to want. Native functionality rather than screens that ignore the device. iOS live activities. NFC. Wearables. Spatial design. Conversational design and AI. Motion treated as communication rather than decoration. Tactile interaction that goes beyond the visual.

None of that was on anyone's roadmap. That was the point. A vision that only contains things already agreed is a plan, and the team had plenty of plans.

Made "what if" part of the job description.

A designer's job is not to work through Jira tickets. It is to ask what if, and occasionally to break the brand guidelines to find out what is on the other side of them.

That had to be taught, not announced. It meant running explorations with no ticket attached, defending them in front of executives, and showing 139 people what it looks like when a designer argues for something nobody asked for.

Three design visions in under three years.

Encapsulated in a "taking personal to the next level" principle, exploring expressive movement, sensory brand experience, intuitive gestures, and depth and scale. I ran executive workshops on each so the ambition was tied to business objectives rather than sitting decoratively beside them.

Turned a rebrand into a digital design language.

The rebrand arrived as new colours, a new logo, new typography and new movement. None of that answers how a brand behaves on a screen.

So I wrote the design theories that translated it. How the logo moves. What motion means as communication. What sensory means for a telco. Whether the brand makes a sound, and what it should be. Then the technical definitions underneath: the spacing model, how the logo is used, how the colours are applied, how movement works. A rebrand hands you assets. Someone has to decide how they behave, and at this scale that decision has to be written down or it does not survive contact with 139 designers.

What I did for the leadership

Rebuilt the leadership layer around five capabilities.

Five core competencies at the centre of design operations: user research, service design, accessibility, product design and content design, with a capabilities leadership layer above the tribe and squad split. The point was decision rights: leaders could make calls about the organisation of design, not just their own squad.

The first restructure moved people between teams and left the real problem intact: above every designer sat a layer whose roles had never required design judgement, so people not equipped to evaluate it kept approving or blocking work. The second version added the Principal layer inside the leadership group, accountable for the vision and for demonstrating what good looked like. You cannot reorganise your way out of a capability gap.

Created the Principal Product Designer profession.

A leadership layer that did not exist anywhere in the company. Piloted with one, expanded to two, focused on pushing the boundaries of the digital products and defining what quality means at tribe level. Principals mentor designers and hold the quality of UI and interaction, which frees Heads of Design to bridge strategy and execution rather than doing both badly. It is probably the best decision I made for the company, and it is the main reason the ambition did not leave when I did.

Defined what a product designer actually is here.

Sixty designers, and no shared understanding of the difference between a UX designer, a UI designer and a product designer. That is not a semantic problem. It is a hiring problem, a levelling problem and a quality problem at once. I wrote the definition, taught it and mentored against it. By the time I left, the profession was 139 people with one description of the job.

Took the design system under my leadership and rebuilt it.

Loop moved into my leadership, under operations, focused on speed and documentation. I rebuilt it from the ground up and built the team around it, including the engineers.

Then we measured it honestly. Building a carousel with Loop took fifteen minutes and 63 lines of code. Without it, forty-five minutes, 175 lines, and the engineer gave up. That comparison did more for adoption than any amount of advocacy, because it argued in terms engineering already cared about. Adoption went from roughly 12% to 80%, with 90% token utilisation.

Built the training system, and the people to carry it.

Structured learning paths mapped to career progression, so training became a route to promotion rather than an extra. Learning Days with hands-on workshops. 220 designers trained in variables and advanced Figma prototyping. Sixteen interns over two years, trained alongside the head of content design.

Community engagement went from 22% to 80%. That number is usually read as a training statistic. It is not. It is what happens when people are given something to aim at.

Changed how engineering saw design.

A strategic partnership with Figma, DevMode to cut handoff friction, cross-functional syncs, and a drop-in session where the design system team and engineers made decisions on live projects together.

Scaling the capability of design

Design team, 60 → 139

Designers when I joined

Hires who joined after

Sixty designers, no definition of the job. 139, with a playbook and career progression guidelines.

One year later

Look at EE's homepage today. The spacing model, the way the logo behaves, how colour is applied, how movement is used: those are the definitions I wrote, still directing how the company designs a year after I stopped working there.

Visions are easy to present and hard to make durable. Most are admired once and filed.

The documents were never really the deliverable. The definitions inside them were, written to be picked up and used by people I would never meet. That is the part of this job I care most about getting right.

The outcome

80%

80%

80%

80%

engagement in the design community, up from 22%.

40%

40%

40%

40%

fewer UI bugs

95%

95%

95%

95%

accessibility score

£2.3M

£2.3M

£2.3M

£2.3M

found in hardcoded components, later saved when moved to Loop Design system

220

220

220

220

designers trained on EE and BT Group

The work is in the live product. A year after I left, EE's homepage runs on the design language, spacing model, and the motion rules I defined. Nothing about that needed me to be there, which was the design goal.

The profession changed shape. Sixty designers with no definition became 139, with a job description, a career ladder, and a Principal layer holding quality above the squads.

The system got used. Adoption went from roughly 12% to 80%, token utilisation to 90%, and the app's UI bug rate fell 40% from a starting point of 60%.

And design became legible to the business. The people strategy moved 20% of design into managed service providers, freeing permanent roles for the strategic work.

In retrospect

I wanted the visions built while I was there, and for a while I read the fact that they were not as a failure. It was not. Vision work at that scale runs on a slower clock than product work, and the job is to make the ideas durable enough to outlive your own impatience. Judging your work by whether it ships during your tenure is the wrong measure, and it took leaving to see that.

Saying "twice" out loud about the restructure is the point. At that scale you are not redrawing a chart, you are reorganising people's working lives, and getting it wrong once teaches you to consult far more widely before the second attempt.

At 139 designers, my medium stopped being Figma. It became the deck and the room. I have hundreds of presentations from those two years, and it took me a while to stop treating that as overhead and start treating it as the job. A design decision that cannot be explained to an executive does not get made.

NEXT CASE

Cazoo