← Back to Clara
AgentClara
Reduce decision complexity
Finds confusing choices and improves grouping, naming and timing without blindly removing useful controls.
decision-complexity-hick-hyman-analysis
See how it helpsWhen is it useful?
Use when people hesitate, select the wrong action or lose context in a multi-step interface.
What does it help you do?
- Distinguishes meaningful options from duplicate, premature or already-known decisions.
- Groups choices around recognizable tasks and separates rare consequential actions.
- Balances expert shortcuts with understandable paths for less familiar users.
Try asking this
Review our project-creation flow and propose clearer choices without turning one coherent task into an unnecessary wizard. Group choices by the project users want to create. Keep advanced settings available, and compare a newcomer's path with an expert's repeated task.
Detailed documentation
Read the skill’s original instructions, examples and checks in their authored language.
Show less
Decision Complexity & Hick–Hyman Analysis
Decision time and error rates grow not merely from the raw number of buttons on a screen, but from ambiguous, overlapping, poorly grouped, or high-risk choices that force the user to stop and deliberate.
Nuanced Application of the Hick–Hyman Law
- Do NOT apply Hick's Law as a simplistic "fewer buttons everywhere" rule:
- A well-organized 12-item navigation sidebar with clear, mutually exclusive categories is faster to scan than a 3-item mystery menu where the user has to guess which bucket hides their tool.
- Splitting a cohesive 1-screen task into a 7-step wizard just to show "one question per screen" increases memory load, clicks, and context loss.
- Focus on what actually drives decision latency ($H = \sum p_i \log_2(1/p_i + 1)$):
- Option discriminability: Can the user tell in 1 second how Option A differs from Option B?
- Familiarity & frequency: High-frequency expert actions (
Save,Run,Filter) have near-zero decision cost when placed consistently; rare, high-consequence choices (Delete workspace,Choose billing model) require clear framing. - Categorization structure: Grouping 15 flat items into 3–4 meaningful task-oriented clusters turns a linear scan into a fast two-step recognition.
Workflow
- Spot and eliminate false or premature decisions:
- Flag places where the UI asks the user to choose something the system already knows or that 95% of users never need to change before starting.
- Flag duplicate ways to trigger the exact same action on the same screen with slightly different names (e.g.,
"Create Task","New Work Item", and"+ Add"side by side).
- Make remaining choices mutually exclusive and concrete:
- When a user must choose between modes, roles, or templates, show the concrete consequence of each choice in plain language (1-line subtitle or preview) rather than abstract internal type names.
- Reduce consequence anxiety:
- Users hesitate when they cannot tell whether a button saves a draft, immediately charges a card, or overwrites live production state. Make reversibility and blast radius explicit on the action label itself (
"Save draft"vs."Publish to live workspace").
- Users hesitate when they cannot tell whether a button saves a draft, immediately charges a card, or overwrites live production state. Make reversibility and blast radius explicit on the action label itself (
Included files 1
SKILL.md
