Solution Space

Solution space describes how many valid ways exist to solve the problem in front of you. Narrow space means one obvious correct answer. Wide space means multiple good paths with meaningful tradeoffs in craft, speed, scalability, or brand alignment. The concept exists so designers stop measuring task difficulty by file size or hours spent and start measuring it by ambiguity. Effort dial settings only pay off when the solution space is wide enough to reward deeper thinking.

A rename task has narrow solution space. Five headline variations has medium space. Choosing between two design system architectures has very wide space because both approaches can work but create different downstream costs. The width determines whether extra reasoning actually helps or just slows you down.

It is not the same as task size. A 400 screen Figma file can still have narrow solution space if every decision is already constrained by the existing system. It is also not the same as project importance. You can have a high stakes but narrow task that still does not need max effort. The signal is ambiguity, not gravity.

Designers regularly confuse solution space with effort required. They see a big project and immediately set max effort without checking how many real decisions remain open. The result is overthought obvious work and underthought ambiguous work. The dial only makes sense when you read the space first.

The product team at Raycast in 2025 used solution space language explicitly in their AI workflow. Renaming every command in their palette was narrow space so they stayed on low effort and finished in minutes. Deciding how their new agentic first run should expose power surfaces was wide space. They ran xhigh and let the model map four different mental model approaches against their existing keyboard layer and streaming surface patterns. The extra reasoning surfaced a hybrid pattern none of the humans had considered.

A design director at Notion ran the same test on their empty state archetypes project. The team needed a new system for progressive disclosure across ten different use cases. The solution space was wide because each archetype carried different tradeoffs in perceived performance and delight. Max effort delivered a structured rubric with clear recommendations. Medium effort on the same task had simply listed ten variations without weighing anything.

Reach for higher effort settings only when the solution space is wide. Narrow space tasks earn low or medium because extra reasoning adds nothing but latency. Wide space tasks with high cost of being wrong earn xhigh, max, or ultracode because the model can actually map tradeoffs you might miss. The skill is diagnosing the space before you touch the dial.

The tradeoff is calibration. Read the space wrong and you either overpay for simple work or underthink risky work. Most teams improve this skill by running quick audits after the fact. Was the output over hedged? The space was probably narrow. Did we miss an obvious tradeoff? The space was wider than we diagnosed.

Solution space is the question that should precede every effort command. How many good answers exist here and how different are their consequences? Answer that first and the dial almost chooses itself.