Most briefs describe what needs to ship. The real work is deciding what needs to be true two years later.

The pattern in my best work is the same each time. Someone hands me a narrow brief, I work out where the thing is actually headed, and I build the version that keeps paying off after the brief is closed.

At Google the brief was one designer to fix some tools. Instead, I built a UX org from one to 32 that can shape the roadmap, strategy, and vision. At Amazon the brief was a recommendation widget. I built a modular platform that now runs on 80% of searches and added over $1B in annual sales. With the IRC the brief was one crisis site for Ukraine. I shipped it, then turned it into a system that launches a new site in under a week and has reached 85M+ people across 20+ countries.

Philosophy

Three kinds of brief, one rule. Whether I’m handed a screen, a team, or a new technology, I work out what has to be true in two years and build for that.

  1. 01

    Design

    A design that doesn’t solve a real user or business problem is art. I like art. It isn’t what I’m paid for.

    Results are the most powerful form of influence
    Skeptical teams came around to test results in their own category, not a pitch deck. Amazon
    Every design is a hypothesis
    All of it tested, two weeks at a time, and the data outranked my taste. I was right that shoppers wanted someone to trust. I was wrong about who. Amazon
    Systems before screens
    A modular template or a shared design system makes the next version almost free. Amazon’s ran on 80% of searches; the IRC’s launches a new site in under a week. Amazon · IRC
    Familiar beats clever
    In a crisis or a security console, learning time is risk. Icons before words, conventions before inventions. In testing, refugees said the IRC’s icons looked like the ones already on their phones. IRC
  2. 02

    Leadership

    Culture is the operating system. If the how is broken, the output will be too. I split the job in two and refuse to do only one half.

    Develop the team

    Make it safe to be bad at something in public
    AI adoption went from 34% to 100% in three quarters, and the habit underneath it outlasted the program. Google AI
    Empower people and don’t micromanage
    The biggest idea in my AI program came from a UX engineer on my team. My job was to recognize it and make it real. Google AI
    Develop individuals and support careers
    18 promotions. 19 of 21 direct reports still on the team. Google
    Coach, give feedback, listen
    Two-hour open critique blocks twice a week, bookable in 30-minute slots.

    Get things done

    Set the vision and strategy with the team, not for it
    Annual workshops where the whole org writes its values, mission and OKRs. "Make it easy to protect and defend Alphabet" is still the line we use. Google
    Be decisive, and give every no a reason
    A backlog scored on user value against effort means weak ideas can be turned down with the reasoning attached. Google AI
    Make the org legible
    Written engagement models, co-driven projects, short spikes, office hours, so partners know how to get UX and what they’ll get. Google
    Keep the technical skills to do the work
    I lead a team of managers and UX engineers and I still ship. AI Lab
  3. 03

    AI

    When Google pushed to put AI in everything, I wrote the strategy first. AI has to solve a real problem for the person using it, or it’s a hammer seeking a nail.

    Should we, before can we
    The scarce thing is attention, not engineering. The designer’s job shifts to curation: fewer, better things. Google AI
    Code is the material
    Static mocks hide the awkward state; a working prototype can’t. Google AI
    Shed the process every few months
    Model capabilities move faster than playbooks. I retired my own framework once it had run its course. Google AI
    Demos over memos
    Prototypes replaced specs. Small, frequent critiques replaced big-room reviews. Google AI

Off the clock

When I’m not working, I’m usually navigating the wonderful chaos of family life in Chicago with my wife, our three children, and our dog, Cooper. I also love gardening, cooking, and tumbling down a good internet rabbit hole. If I’m lucky, I might even sneak in a round of golf.