Tooltip guidance
Low interruption and easy to reuse, but less discoverable and harder to retain during a complex task.
STEELSERIES GG · SOFTWARE UX · 2024–PRESENT
Designing complex gaming controls for a 5M+ user platform.
I shape software experiences across SteelSeries GG, from product discovery and data analysis to interaction design, validation, engineering handoff, and launch.
5M+monthly active users
100K+data points analysed
110+external research participants
20+in-depth interviews
PROJECT OVERVIEW
A mature ecosystem with very different user needs.
GG serves players who want a quick setup and enthusiasts who expect precise control. The product challenge is to make advanced settings understandable without slowing expert users or creating one-off patterns for every device.
- Role
- Senior Product Designer
- Scope
- Engine, Sonar, Moments, 3D Aim Trainer, and new device programmes
- Partners
- Product, engineering, firmware, hardware, research, and regional teams
- Ownership
- Discovery, product definition, software UX, validation, design system, handoff, and design QA
My role
I frame the decision with product and engineering, triangulate behavioural data with research, prototype the riskiest interaction, and review implementation against the intended states and success criteria.
01 · DISCOVER
Triangulating behaviour, feedback, and product data.
I do not rely on one research method. I combine large-scale behavioural evidence with interviews and usability work, then trace each design decision back to the signal that caused it.
100K+
records analysed across product and research sources
Redash and Snowflake revealed usage patterns. Airtable and Dovetail organised feedback and research evidence. Python and SQL helped clean, join, and compare the data before synthesis.
RedashSnowflakeAirtableDovetailPythonSQLDiscordNPSUseberry
Recurring questions and language from active players
Patterns behind positive and negative experiences
Evidence tagged by product, topic, and severity
Preference and behaviour patterns across participants
Internal metrics, participant identities, account details, links, project codes, and verbatim feedback have been obscured. The visuals show method and evidence structure only.
02 · DEFINE
Turning evidence into a product decision.
For advanced keyboard configuration, research showed that first-time users could see the controls but did not understand the underlying concepts. The problem was information timing, not simply terminology.
- SignalPlayers misinterpreted advanced settings.
Interviews, community feedback, and usability sessions surfaced the same comprehension gap.
- Root causeTechnical controls appeared before the mental model.
More copy alone would increase density without helping users form the right concept.
- HypothesisProgressive guidance could teach without slowing experts.
Contextual education would appear only when the user needed it.
- ExperimentTooltip guidance versus expandable guidance.
Useberry compared comprehension, confidence, task success, and completion time.
- DecisionUse evidence to select the interaction pattern.
The decision gate compares comprehension, confidence, task success and time across both patterns; I document where each pattern serves novices or experienced players. The study does not establish a quantified live-product lift.
Alignment as a product tool
I facilitated FigJam workshops with Product, Design, Engineering, Hardware, and Firmware. We compared user impact, technical complexity, dependencies, and product risk before committing to scope.
The output became a focused problem statement, prioritised opportunities, success criteria, and a validation plan.
FigJamOpportunity mappingRoot cause mappingPRD
03 · DEVELOP
Exploring the interaction before polishing the screen.
I move from low-risk structure to high-fidelity behaviour. For important decisions, I compare multiple patterns, document trade-offs, and test the smallest prototype that can answer the open question.
Expandable guidance
More persistent context and room for examples, with a larger layout and interaction cost.
Useberry A/B study
Measured task success, comprehension, confidence, and time. Qualitative feedback explained the score differences.
Behaviour, states, and realistic task context
Task behaviour and attention patterns compared across options
04 · DELIVER
Owning quality beyond the Figma file.
The design process continues through planning, engineering handoff, implementation review, alpha and beta testing, and design QA. I document responsive behaviour, component states, edge cases, and the constraints that affect the final experience.
Requirements and scope
Connect the product brief to user flows, acceptance criteria, and Jira tickets.
System-ready handoff
Specify SteelForge components, behaviour, states, and edge cases with engineering.
Implementation review
Compare the build with the intended interaction and report gaps with clear reproduction steps.
Release learning
Review alpha, beta, and post-launch signals, then carry findings into the next iteration.
Requirements translated into buildable work
Dependencies and programme timing aligned across teams
Findings connected back to tasks and decisions
Internal and player feedback before release
DESIGN SYSTEM
Building reusable product decisions into SteelForge.
I design within SteelSeries’ existing system and contribute patterns that can scale across products. Influenced by Material Design 3 principles, the work covers tokens, components, states, interaction rules, documentation, accessibility, and design QA.
Foundations
Typography, colour, spacing, elevation, and motion rules.
Components
Sliders, switches, selects, tabs, chips, buttons, tooltips, backdrops, and snackbars.
Behaviour
Loading, disabled, warning, error, confirmation, and recovery states.
Delivery
Shared specifications that improve reuse, implementation speed, maintainability, and cross-product consistency.
GG SOFTWARE ECOSYSTEM
One design practice across different player journeys.
My work spans device configuration, audio control, gameplay capture, and training. The interface changes by context, while shared navigation, interaction patterns, and system rules keep the ecosystem coherent.
Engine
Device setup, profiles, advanced configuration, firmware states, and configuration education.
Sonar
Channel mixing, presets, equalisation, and audio processing across game, chat, and microphone paths.
Moments
Gameplay capture, clip editing, audio control, and sharing within a focused creator journey.
3D Aim Trainer
Training discovery, practice activities, and performance feedback within the wider GG experience.
05 · MEASURE
Connecting usability outcomes to product health.
I define leading and lagging indicators before launch. This lets the team evaluate whether a design improved understanding in the test and whether that improvement translated into healthier product behaviour after release.
Verified scale
5M+monthly active users across GG
The work operates inside a mature platform where consistency and release quality matter at scale.
Research coverage
110+external participants
Supported by 20+ interviews and more than 100,000 product and research data points.
Measurement plan
NPS and support signals
Track sentiment themes and support contacts for the affected interaction after release.
Adoption and daily use
Compare feature activation, repeat use and task completion against a defined baseline.
Leading indicators
Task success · comprehension · confidence · completion time · error rate
Lagging indicators
Feature adoption · DAU · retention · NPS · support volume · community sentiment
Scale and research coverage describe the platform and study programme. Product outcomes for individual releases are confidential or not attributable here; this section therefore shows how I would evaluate them, without claiming a quantified lift.
AI-ENABLED PRODUCT DEVELOPMENT
Extending the design system into an AI workflow.
I built a Codex-driven workflow that converts confirmed meeting decisions and structured requirements into editable Figma proposals using reusable SteelForge rules. The AI drafts component choices and states; I check requirement coverage, interaction logic, accessibility and edge cases before any team handoff. The human review step determines what is usable.
Confirmed decisionsStructured requirementsSteelForge rulesEditable Figma draftHuman design QA
The demo shows how I structure decisions, apply reusable design-system knowledge, generate an editable proposal, and keep human judgement in the review loop.
REFLECTION
Senior design means owning the decision, the system, and the shipped result.
My contribution goes beyond producing screens. I connect evidence to product scope, align teams around trade-offs, design within technical constraints, validate the interaction, and carry quality through implementation and measurement.