Tag: research

  • Error Messages in Practice

    Error Messages in Practice

    Consistency is kind, but clarity wins when the two disagree. Contrast is not a style choice; it is whether people can read the page. We removed three options and support tickets dropped by a third. Lessons Users scan, so headings carry more weight than paragraphs. We tested with five people and found the same problem…

  • Form Design That Actually Help Users

    Form Design That Actually Help Users

    We tested with five people and found the same problem four times. Users scan, so headings carry more weight than paragraphs. Dark mode is a second palette, not an inverted first one. Lessons An error message should blame the system, not the person. Whitespace is not empty; it is doing the grouping for you. Motion…

  • Design Tokens in Practice

    Design Tokens in Practice

    The prototype answered in a day what the meeting could not in a month. Consistency is kind, but clarity wins when the two disagree. Type scale decisions ripple into every component downstream. What didn’t Dark mode is a second palette, not an inverted first one. An error message should blame the system, not the person.…

  • Designing Better Icon Sets

    Designing Better Icon Sets

    A design system is a set of decisions you no longer have to argue about. The prototype answered in a day what the meeting could not in a month. Dark mode is a second palette, not an inverted first one. Lessons We removed three options and support tickets dropped by a third. Whitespace is not…

  • Common Navigation Mistakes

    Common Navigation Mistakes

    Good defaults are the design most users will ever see. The prototype answered in a day what the meeting could not in a month. We tested with five people and found the same problem four times. Going further An error message should blame the system, not the person. Contrast is not a style choice; it…

  • Rethinking Data Tables

    Rethinking Data Tables

    Users scan, so headings carry more weight than paragraphs. Consistency is kind, but clarity wins when the two disagree. Whitespace is not empty; it is doing the grouping for you. What worked An error message should blame the system, not the person. Contrast is not a style choice; it is whether people can read the…

  • Designing Better Design Tokens

    Designing Better Design Tokens

    A design system is a set of decisions you no longer have to argue about. Whitespace is not empty; it is doing the grouping for you. Labels above fields beat clever placeholders every time. The short version Whitespace is not empty; it is doing the grouping for you. Users scan, so headings carry more weight…

  • Rethinking Typography

    Rethinking Typography

    Labels above fields beat clever placeholders every time. Type scale decisions ripple into every component downstream. Whitespace is not empty; it is doing the grouping for you. What worked Users scan, so headings carry more weight than paragraphs. Motion should explain a change, not decorate it. A design system is a set of decisions you…

  • How We Rebuilt Our Design Tokens

    How We Rebuilt Our Design Tokens

    We removed three options and support tickets dropped by a third. Good defaults are the design most users will ever see. Whitespace is not empty; it is doing the grouping for you. Getting started Users scan, so headings carry more weight than paragraphs. Consistency is kind, but clarity wins when the two disagree. The prototype…

  • Common Design Tokens Mistakes

    Common Design Tokens Mistakes

    An error message should blame the system, not the person. Motion should explain a change, not decorate it. The prototype answered in a day what the meeting could not in a month. What worked Good defaults are the design most users will ever see. We removed three options and support tickets dropped by a third.…

  • Rethinking Onboarding

    Rethinking Onboarding

    The prototype answered in a day what the meeting could not in a month. Labels above fields beat clever placeholders every time. We tested with five people and found the same problem four times. Lessons Type scale decisions ripple into every component downstream. Motion should explain a change, not decorate it. A design system is…

  • Common Icon Sets Mistakes

    Common Icon Sets Mistakes

    Type scale decisions ripple into every component downstream. Motion should explain a change, not decorate it. Consistency is kind, but clarity wins when the two disagree. Trade-offs We tested with five people and found the same problem four times. We removed three options and support tickets dropped by a third. A design system is a…