Case study

Unity Technologies

Three initiatives to modernize developer documentation had stalled before mine. Each had been scoped as a content problem.

Role
Senior UX Designer, documentation experience
Timeline
2021, 1.5 years
Team
Senior Product Manager, Engineering Lead, Tech Program Manager, Director of Content and Documentation, Manager of Content Ops
Scope
3M+ pages across 73+ portals

What I shipped

Stakeholder interviews, competitive analysis, IA redesign, workshops, prototypes, executive presentations, final specs, and asset production, run cross-functionally with a Senior Product Manager, Engineering Lead, Tech Program Manager, Director of Content and Documentation, and Manager of Content Ops.

Unity Cloud documentation today: a walkthrough of the platform this infrastructure work made possible.
Doc Content Info Experience Initiative org chart: Web Platform, Web Experience, and Docs Systems and Portals teams, each with its diamond leads and design-and-engineering pods.
The full cross-functional team behind the Doc Content Info Experience initiative.
The Diamond Team view: the Docs Systems and Portals team paired with the Docs Content Ops team around a shared X-Team mission to deliver a smoother documentation experience.
The core “diamond”: product, design, engineering, and content ops, working as one team.

The situation on arrival

This was the fourth attempt. Three prior initiatives had launched without dedicated design, research, and product resourcing, and each had stalled. The documentation served 62,000 daily active users, so every week the fragmentation went unaddressed was a week it compounded at scale.

Developers could not find what they needed, so the answer was more content, better written, published faster. Three teams had already worked from that brief.

What the evidence said

The platform was not failing through lack of content. It was failing through infrastructure. The search index was built from a manually maintained Word document, and one stakeholder described the experience as “something from a different era of computing entirely.” Three independent lines of evidence, from user surveys, stakeholder interviews, and competitive analysis, all pointed to the same structural truth.

Challenges slide: disjointed end-user experience, unclear processes and collaboration, no data reporting, lack of interoperability, lack of variety of learning styles
The five challenge categories surfaced through research, going into the brief.

What was actually true

I started from the UX Research team's prior study, a 1,400-participant effort run the year before I joined, and used it to keep underrepresented learning styles in scope. On top of that I ran internal subject matter expert interviews, a competitive analysis across 12 documentation platforms including Stripe, AWS, and GitHub, and journey mapping against the points where developers gave up or filed a ticket.

The search index was built from a Word document that someone maintained by hand. Seven pods of five to seven writers each had independently built 73+ sub-sites, with no shared navigation and no common brand. Ownership of content was unclear and siloed.

The platform was not failing on content volume. It was failing on infrastructure. Three prior attempts had scoped it as a publishing problem when it was a product problem, which is why each one stalled short of the resourcing it needed. Naming that difference is what secured the resourcing this time, and my selection ran through four management teams and the Director of Product Design.

Activities & outputs

A deeply collaborative process, spanning discovery, synthesis, ideation, and delivery.

  1. 01

    Stakeholder interviews & alignment workshops

    Building shared understanding across content, product, and engineering leadership

    Conducted 12 stakeholder interviews across content, product, and engineering. Mapped misalignments in ownership, priorities, and success metrics.

    Workshop sticky notes: lack of user focus, no unified search, fragmented content, multiple websites, no unified learning paths
    Workshop synthesis: themes and problems clustered from stakeholder input

  2. 02

    Competitive analysis across 12 documentation platforms

    Benchmarking against Stripe, AWS, and GitHub developer docs

    Analysed 12 leading documentation platforms including Stripe, AWS, GitHub, Twilio, and Atlassian. Built a benchmarking matrix used to drive design decisions.

    Competitive analysis grid comparing documentation sites across Unity, Autodesk, Adobe, Microsoft, Salesforce, and Unreal
    Competitive analysis takeaways: clear IA, flexible navigation, dynamic content and support

  3. 03

    User journey mapping & pain-point synthesis

    Mapping where developers got stuck, gave up, or filed support tickets

    Mapped 47 distinct developer journeys across 6 Unity product verticals. Identified 8 primary failure points where users abandoned documentation.

    Holistic problem analysis workshop board: needs, pain points, and opportunities
    Persona: Technical Director at a mid-market game studio, goals and pain points around documentation evaluation
    Persona: Senior Environment Artist, goals and pain points around documentation quality

  4. 04

    Information architecture redesign & content audit

    Redesigning the IA for a documentation platform spanning 3M+ pages

    The work supported a documentation experience spanning 3M+ pages across 73 portals. Proposed a 4-tier taxonomy that consolidated content into a single coherent structure. Validated with content team leads and engineering.

    05

    Iterative prototyping with usability testing

    Validated search patterns, navigation models, and content hierarchy

    Built 6 iterations of the navigation and search prototype over 8 weeks. Ran 3 rounds of usability testing with Unity developers. Each round reduced task completion time by 34%.

  5. 06

    Executive presentations & final design handoff

    Securing stakeholder buy-in and handing off production-ready specs

    Presented the final design system to the CTO, Director of UX Design, and technical writing teams in a company-wide town hall. Delivered production-ready specs to engineering within 2 weeks.

Forging our principles

Before any pixels, the team needed principles. These three became the north star for every decision that followed, from information architecture down to interaction and information density.

Design principles slide: Predictive, Adaptive, Integrated
The slide the team carried into every review.

Principle 1

Predictive

Understanding what users need or may need, and how to surface that information before they know to ask for it.

Principle 2

Adaptive

Users learn in more than one way, so the documentation had to offer real alternatives, whether video, interactive examples, or written reference.

Principle 3

Integrated

Meeting developers where they already are, in the creative or build process, with right-time documentation that fits the workflow instead of sitting alongside it.

“Every decision we made had to pass a simple test: does this make it faster and easier for a developer to solve their problem?”
Christopher Miller, Senior UX Designer
“It's great that you're doing this, because it'll help the group come together and start to gain momentum.”
Mary Luther, Director, Product Design, Unity
“Running a workshop with the Docs Leads was a crucial step to ensure everyone was on the same page.”
Brian Coughlin, Senior Product Manager, Unity

The Outcome

What changed

Navigation and search were the unlocks, so the work went there first: a redesigned information architecture for a platform spanning 3M+ pages, a real search system to replace the hand-maintained index, and consistent UX and documentation process across what had been 73 separate properties.

Three principles held the decisions together. Anticipate what a developer needs before they know to ask. Support more than one way of learning, through video, interactive examples, and written reference. Meet developers inside their workflow rather than alongside it.

Docs landing page before and after: fragmented single-column layout replaced by a structured hero, getting-started section, and browse-by-product index
The docs landing page, before and after: same URL, restructured around getting started and browse-by-product instead of a flat link list.
73 to 1

73+ independently built portals, each with its own navigation and no shared brand, consolidated into one documentation experience with a single information architecture and one search index.

−34%

34% fewer support tickets in the 90 days after launch: fewer developers hitting a dead end and filing a ticket instead of finding the answer.

+41%

41% increase in session depth: once people found what they needed, they kept going instead of bouncing back out to search again.

Building trust through transparent communication

Context

Unity's documentation initiative was launching for the fourth time after three failed attempts. Technical writers, content teams, and developers had been burned before: skepticism was high, and trust needed rebuilding.

What made it work

I framed the sessions around Unity's core values, "best ideas win" and "we're in it together." High-fidelity prototypes went up alongside the research that informed them, design decisions were explained rather than announced, and the question was always open: "This is where we're at. What are your thoughts?" The goal was to rebuild trust by proving we were listening.

My approach

I treated town halls as collaborative checkpoints, not presentations. Every quarter I presented current state, design prototypes, and research findings to an audience spanning individual contributors up to the CTO and VP of Product, across disciplines. These were working sessions designed to create transparency and invite real-time feedback.

The result

Across six company-wide town halls, feedback from technical writers to the C-suite shaped the roadmap, surfaced cross-functional dependencies early, and turned a legacy of failed initiatives into organizational buy-in. Doing it in the open is also what taught me to navigate skepticism and communicate complex design decisions to non-design audiences.

“Transparency isn't about having all the answers: it's about showing your work, inviting feedback, and proving people's voices matter.”
Christopher Miller, Senior UX Designer
QBR town hall title slide: Docs User Insights
One of six quarterly town halls: Docs User Insights.
Screen recording of a town hall walking through the docs landing page design update, with prototype annotations and a live call audience
Walking the redesign through with the group, live: prototype and rationale together.

Designing for the long game

  1. Search is infrastructure, not a feature. Investing there first unlocked everything downstream.
  2. IA decisions are product decisions. Content structure directly shapes how capable the platform feels.
  3. Cross-functional ownership was the hardest and most important part. Design couldn't solve this alone.
  4. Measuring the right thing matters. Support tickets and session depth told a story page views never could.
  5. Vendor decisions are design decisions. Search platform candidates like Algolia and Coveo carried real tradeoffs between relevance quality, cost, and lock-in risk, and surfacing those tradeoffs alongside the IA work mattered independently of which one the business eventually chose.

Final screens

The shipped documentation home, the Unity Editor product page, and the Game Server Hosting product page, as handed off to the lead engineering team.

The shipped documentation home page: a search-led hero, a getting-started row, and a browse-by-product index.
Documentation home.
The shipped Unity Editor product page: featured resources, learn-about cards, and a sample-project row.
Unity Editor product page.
The shipped Game Server Hosting product page: an overview, popular links, next-steps guides, and related documentation.
Game Server Hosting product page.

Where it stopped

My part ended at final design handoff, with production-ready specifications and executive sign-off across content, product, and engineering leadership. Rollout across the remaining portals continued on the content and engineering side after handoff. The consolidation was staged rather than a single cutover, so portals moved over on their own schedules.

Close collaboration with the Senior Technical Program Manager and the Lead Engineer delivered what three previous initiatives could not. The design system the team carried forward, internally called "Guardian," was a direct outgrowth of the IA and search work here.

The redesigned Unity Cloud Documentation site shown in a laptop browser frame: a left-hand navigation tree, an overview page, and cards for Asset Manager, DevOps, and getting started.
The documentation experience this work rebuilt, shown as it shipped.

Here's what they say

Words from my teammates and leaders.

More work