Product decisions

Does your product need a node editor?

Sometimes a list is enough. Sometimes the connections are the product. How to tell the difference before you build.

By FlowField26 Sept 20263 min read

A sequence tells you what comes next. A canvas can also show where the work branches, rejoins, or depends on something else.

01 / Product decisions

Start with what people need to see.

A customer sends a request. Someone checks it. Someone approves it. If the process always follows those three steps, a list might explain it perfectly. Drawing lines between three boxes does not automatically make a better product.

Now add different approval paths, reusable checks, and work that can happen in parallel. The question changes. People need to understand how the pieces relate, not just which task comes next. That is where we start considering a canvas.

02 / Product decisions

Look for decisions your users own.

A canvas becomes useful when people need to shape the process themselves: change a condition, connect an existing flow, or compare two possible routes. Those are meaningful product actions. Moving a box around is only useful if it helps people do that work.

Ask who changes the workflow, how often it changes, and what they need to know before changing it. If the answer is always a developer, once a year, a visual builder may be an expensive interface for a setting nobody touches.

03 / Product decisions

Make each connection mean something.

Before designing a node, finish this sentence: this line means… It could mean that one task starts after another, that information passes between two steps, or that a condition has been met. Pick a meaning people can learn and use it consistently.

In our work on Alentra, the workflow brings wallet identities into a reusable setup for sessions. In ProseID, the canvas connects legal flows and integrations. The boxes serve different purposes because the products solve different problems. A useful editor starts with those differences.

See Alentra and ProseID

04 / Product decisions

Design for the first mistake.

Imagine someone connects the wrong two steps. Can they understand what happened, undo it, and try again? Can they spot an unfinished branch before running the workflow? These questions matter at least as much as how smoothly a node drags.

We want the editor to make the safe path easy: useful defaults, clear connection rules, and errors explained beside the work that needs attention. A person should not need to understand the underlying software to recover from a mistake in the interface.

05 / Product decisions

Test the idea before building the editor.

Take one real task and sketch it twice: once as a straightforward form or list, and once as a canvas. Ask someone who does the work to explain each version back to you. Then ask them to change something.

Listen for where they hesitate. If the canvas helps them see consequences and make decisions, you have a reason to explore it. If they spend their time finding the next box, simplify. We love node editors. We still want every one we build to earn its place.

Building products3 min read

A new product starts with one real journey.

Before the feature list gets longer, follow one person from the first action to a useful result.

24 Sept 2026 Read article

All articles

Svelte Flow

Let’s talk