Excalidraw: let the first action be drawing

A tool on the way to an illustration

In his 2020 retrospective, vjeux described Excalidraw's beginnings as a small tool for hand-drawn-looking diagrams. He protected a quick start, resisting title-entry and login steps before drawing. He also acknowledged the advantages of experience, an existing audience, and personal connections. One delightful detail is that a tool-making detour eventually supplied the illustrations he had wanted for a blog post.

This is a short reading of the original account. The design suggestions below are Shilu's own observations, not the author's instructions or a promise of repeatable growth.

What does the visitor want to do first?

Many first screens introduce a product before allowing its central task. A drawing tool suggests a simpler question: if the visitor already knows what to draw, could the canvas come first? The equivalent for a compressor is selecting an image; for a time-zone planner, entering cities.

The answer depends on the task. A purchase, a sensitive action, or an invitation may require information before proceeding. But an introductory screen should earn its place. Write down every step between arrival and the first useful result. For each one, ask whether it helps the visitor or collects something the product would like to have.

A simple screen still needs careful work

A nearly empty canvas can conceal difficult questions about saving, undo, accidental taps, and small screens. Our recommendation is to consider an easy beginning and a safe ending together. The first action should be apparent, mistakes should be recoverable, and the person leaving should know whether their work has been saved.

Sharing adds another responsibility. A fast start should not leave people uncertain about who can see their work. A quiet interface can still make essential status visible. Simplicity is a choice about what deserves attention, not an excuse to remove information that users need.

Try subtracting one obstacle

Ask a first-time visitor to complete one task on your own homepage. Observe the first half minute without explaining the interface. Record where they click, pause, or ask a question. Remove the earliest obstacle and repeat the observation before changing many other things.

This small experiment will not predict commercial success. It can, however, reveal whether the product lets someone approach their own goal. That is the part of the story we find useful: care for the first step, the final step, and each interruption between them.