AgentSkillsCN

friction-logging-and-ux-reviews

在产品开发过程中，通过角色扮演特定用户画像，深入挖掘“易碎边缘”，从而系统性地审计用户体验。适用于重大发布前的开发阶段，或作为定期审计，持续为产品品质树立高标准。

SKILL.md
--- frontmatter
name: friction-logging-and-ux-reviews
description: A systematic method for auditing user experiences by role-playing specific personas to identify "broken edges." Use this during product development, before major launches, or as a recurring audit to maintain a high bar for craft.

Friction Logging and UX Reviews

Friction logging is the practice of systematically identifying every hurdle, point of confusion, and "broken edge" in a user flow. By adopting a specific user's mental model and recording a stream-of-consciousness log of the experience, teams move beyond abstract metrics to granular, actionable improvements that compound into a superior product.

The Solo Friction Log Process

Perform this audit individually on a recurring basis (e.g., once a month) or when a new feature is ready for its first end-to-end test.

  1. •Define the Persona: Choose a specific user profile to mental-model. Do not just "test the product"—be "an engineer at a large SaaS company integrating a billing API for the first time."
  2. •State the Goal: Define exactly what the user is trying to achieve (e.g., "Set up a subscription with a 30-day trial").
  3. •Start the Stream of Consciousness: Open a blank document. Record every thought, frustration, and moment of delight as you move through the flow.
  4. •Note the Details:
    • •Context switching: "I had to leave the dashboard to find a secret key in the docs."
    • •Error handling: "The error message was generic; it didn't tell me what field was missing."
    • •Latency: "This page took 3 seconds to load, making me wonder if it crashed."
  5. •Include Praise: Explicitly mark things that work well. This reinforces high-quality patterns across the team.
  6. •Tag and Distribute: Share the log with the relevant PMs and Engineering Managers. Tag specific owners for "paper cuts" (minor issues) and "blockers."

Group Review: "Walking the Store"

For critical product flows, transition from solo logging to a collaborative review involving cross-functional leadership (Engineering, Product, Design, and Support).

  1. •Select the Flow: Pick a high-stakes path, such as onboarding, checkout, or a complex API integration.
  2. •Shared Issue Log: Create a live document or "crying octopus" (internal feedback) link where attendees can type issues in real-time while the leader walks through the flow.
  3. •The Walkthrough: A PM or Engineer shares their screen and performs the task live. They should talk through their choices and assumptions.
  4. •Debate the "Bar": Discuss identified issues immediately. Ask: "Does this meet our standard of being meticulous? If not, what is the fix?"
  5. •Establish Shared Language: Use these sessions to calibrate the team’s definition of "good" so that future micro-decisions align with the company's craft standards.

Friction Log Template

  • •User Persona: [e.g., First-time developer at a startup]
  • •Task: [e.g., Issuing a refund through the dashboard]
  • •Log:
    • •[00:00] I’m looking for the 'Payments' tab. It was easy to find. (Praise)
    • •[00:01] I clicked 'Refund' but I'm not sure if it worked because there was no loading state. (Friction)
    • •[00:02] Received error code ERR_742. I have no idea what this means. I have to Google it. (Friction)
    • •[00:05] Found the fix. Why didn't the error message just link me to the right doc? (Opportunity)
  • •Summary of Fixes: [List of actionable tickets]

Examples

Example 1: Improving API Error Messages

  • •Context: A developer is integrating an API and makes a syntax error.
  • •Observation: The friction log noted that the developer spent 10 minutes searching documentation to interpret a cryptic error code.
  • •Application: The team added code to the API error handler to detect the specific error and provide a direct URL to the relevant documentation page in the response.
  • •Output: Reduced integration time and increased developer "delight."

Example 2: Optimizing Checkout Flows

  • •Context: A user is checking out on a mobile device.
  • •Observation: During a "Walk the Store" session, the team noticed a user had to click three times to select a saved payment method.
  • •Application: The team prioritized removing clicks and reducing latency in the "payment element."
  • •Output: A measured revenue uplift of 10.5% for users who migrated to the optimized flow.

Common Pitfalls

  • •Testing as a "Power User": Avoid using your internal knowledge to skip steps. If you know where a button is because you built it, you aren't logging real friction. You must maintain the "beginner's mind."
  • •Ignoring "Paper Cuts": Small issues (like a typo or a slightly off-center icon) are often dismissed as "non-blocking." Collectively, they signal a lack of craft. Fix them to prevent "death by a thousand cuts."
  • •Lack of Persona Specificity: Logging friction for a "general user" leads to vague feedback. Be specific (e.g., "A CFO at a Fortune 500 company trying to export a CSV") to uncover high-value nuances.
  • •Making it Punitive: UX reviews can be stressful. If the culture becomes about "catching" mistakes rather than "co-creating" quality, teams will hide friction. Always start with the goal of helping the user.