Case 04 · Kaizen Gaming · Under NDA
Design that survived growth
Seven years at Stoiximan and Betano, growing design from one team into a fifty-five-person function — and building the structure that kept quality repeatable.
- Role
- Lead Designer → Product Design Manager
- When
- Apr 2018 — Apr 2025
- Where
- Kaizen Gaming · Stoiximan / Betano
Challenge
A small design team inside a fast-growing operator, working like a service desk where quality depended on whoever picked up the ticket.
Approach
Standards, critique and career expectations before headcount, then designers embedded in product teams on one shared definition of quality.
Result
A multi-team design function that outlasted me, with evidence as the default and shared patterns lifting quality across products.
The problem
What was actually wrong
Sportsbook and gaming at scale is a hostile environment for design quality: an enormous surface, content that changes by the second, and every market wanting something slightly different.
The craft was there; it did not survive growth. There was no shared standard to critique against, no structure to hire into, and no way for a designer to see what better looked like.
Before
After
My role
Who did what
Mine
- Grew design from a small team into a mature multi-team function.
- Put the standards, critique rituals and career expectations in place that made quality repeatable rather than personal.
- Led design across the core betting and gaming experiences in several markets.
- Mentored designers and design leads — craft, product thinking and influence.
The team’s
- Designers owned their product areas end to end and brought the evidence to critique.
- Product and engineering leadership held the roadmap and the constraints the work had to survive.
- Research and analytics supplied the observation decisions were made against.
Process
How it went
- 01
Name the standard before hiring
Career expectations and a shared definition of quality first, then leads mentored to hold them — otherwise growth just multiplies inconsistency, and the function stops at what one person can review.
- 02
Make critique a ritual
A recurring, structured critique with rules about what is being asked for. Feedback stops depending on who is in the room.
- 03
Move decisions off opinion
Research alongside experimentation, so a disagreement resolves into a test rather than into seniority.
- 04
Share patterns across products
Reusable patterns and shared standards raise quality everywhere at once, instead of product by product.
Before
After
Decisions
What was chosen, and what it cost
The career ladder and standards before headcount
- Over
- Hiring first and formalising later
- Because
- People join a structure or they join a mood. A designer who cannot see what better looks like invents their own definition, or leaves.
- It cost
- Slower hiring at exactly the moment the roadmap wanted more hands.
Continuous observation of the product
- Over
- Research rounds scheduled per project
- Because
- In a product that changes by the second, a study is out of date by the time it is written up.
- It cost
- A commitment with no finish line, defended in every planning cycle.
Reusable patterns across products
- Over
- Letting each team optimise its own surface
- Because
- Local optimisation across many teams produces something that feels like several products.
- It cost
- Teams lose some autonomy, and the system has to earn it back by being faster to use than to ignore.
Outcome
What came of it
Design went from one team to a fifty-five-person function with standards, critique and career expectations — and stayed that way after I left, which is the only real test of whether the structure was ever mine or the organisation's.
Evidence became the default: research and experimentation shaping decisions around outcomes rather than output.
Shared patterns and standards lifted quality, consistency and delivery speed across products rather than inside one.
Lesson
What it cost me to learn
Most of this work lives inside production betting products and cannot be shown as screens. What is portable from seven years here is the operating model, not the artwork.