Evolving design on a trusted tool that millions depend on

Balsamiq Wireframes is used by over 1.4 million designers, product managers, and developers worldwide. They choose it for one reason: it gets out of the way and lets them think through ideas and find the right design.

But a tool that’s been in market for over a decade accumulates friction. Core workflows that made sense in 2010 had developed gaps, and expectations shaped by tools like Figma meant that what once felt fast now felt constrained. We needed to modernize the core experience without breaking the trust and muscle memory of millions of users.

The work was less a visual refresh than surgery on a living product: re-imagining how core interfaces and workflows behave while keeping the tool stable, familiar, and effortless to learn.

|

Commenting worked out in wireframes first, in the tool itself, before engineering committed time

Discovery kept pointing at the same places. Text editing, commenting and review, and canvas navigation were where the tool was actively slowing people down.

I paired the quantitative signals, support tickets and feature requests, with user interviews and contextual observation, and turned them into a prioritized map of usability gaps. The point was to separate what was genuinely costing people time from what they merely wished for. Basic behaviors like WYSIWYG text editing landed in the first pile, and so did commenting and review, the core of how teams collaborate in the tool.

Each gap was a different kind of design problem. Canvas navigation no longer matched how people actually worked. Editing and commenting hadn’t kept pace with what modern editors taught people to expect. And the AI features were a new problem entirely, less about layout than about helping people reach better ideas and get to "done" faster.

I led design from discovery through release: wireframes, visual design, prototyping, and design review with engineering.

The workflow I established kept multiple parallel projects aligned. From the early years, when a small team shipped web and native Mac and Windows versions at once, I maintained the design system—component library, interaction patterns, platform-specific guidelines—that gave us velocity without sacrificing consistency. Wireframed flows and prototypes meant the product team always had a clear picture of the experience we were building toward.

For AI features, my role extended past traditional design. I worked with engineering on the system prompt, rules, and guardrails that governed the model’s behavior, contributed domain knowledge about how LLMs construct HTML to improve generative layout, and helped define the evaluation criteria we reviewed generated wireframes against. The feature went on to convert qualified trial users at 3–4× the rate of those who didn’t engage with it.

Fifteen years on one product is its own kind of design education.