Heuristic Evaluation Report

evaluationmediumAdvanced

TL;DR

Expert interface evaluation based on established usability principles.

What is it

The Heuristic Evaluation Report presents results from an expert audit based on recognized usability principles (like Nielsen's 10 heuristics or Shneiderman's principles). Unlike a usability report, it doesn't require users — it's evaluated by UX experts.

What it is for

  • Identify usability problems without needing to recruit users
  • Quickly evaluate an existing product or prototype
  • Complement usability tests with expert perspective
  • Create a usability baseline before a redesign

Research methods that feed it

Heuristic evaluation (Nielsen)Cognitive WalkthroughWCAG accessibility audit

When to use it

  • When you need a quick evaluation without recruiting users
  • As a step before usability tests (to focus the tests)
  • When auditing an existing product before redesign
  • To verify compliance with usability and accessibility standards

When NOT to use it

  • As a substitute for tests with real users (complement, not replacement)
  • If you don't have evaluators experienced in heuristics
  • To evaluate user needs or behaviors (that requires user research)

How to create it step by step

  1. 1Select heuristics: Choose the framework (Nielsen 10, Shneiderman 8, etc.).
  2. 2Define scope: Which flows or screens to evaluate.
  3. 3Evaluate individually: Each evaluator reviews independently (minimum 3-5 evaluators).
  4. 4Classify problems: For each problem: violated heuristic, severity (1-4), screen, evidence.
  5. 5Consolidate findings: Combine all evaluators' findings, removing duplicates.
  6. 6Prioritize and recommend: Order by severity and provide specific recommendations.

Tips for small teams

  • 3 evaluators capture ~75% of problems — you don't need more
  • Use Nielsen's 10 heuristics as the standard base
  • Combine with 1-2 usability tests to validate the most critical findings
  • Create a reusable evaluation template

Common mistakes

  • Evaluating with a single expert (individual bias)
  • Not classifying by severity (everything seems equally important)
  • Confusing personal preferences with real heuristic problems
  • Not including visual evidence (annotated screenshots)

Contextualized example

Context: Heuristic evaluation of government services portal.

Finding H1 (Visibility of system status): The 5-step form shows no progress indicator. Users don't know which step they're on or how many remain. Severity: 3/4. Recommendation: add progress bar with step names.

Related deliverables

Related methodologies

Free tool by UXR — UX Research Consulting in Chile