Understanding Apdex Score
How we calculate. Apdex = (satisfied + 0.5 × tolerating) ÷ total samples (0 to 1 scale). The form uses the same arithmetic as the worked examples on this page. See our methodology and accuracy policy.
Real-world scenario: A typical Apdex Score case uses satisfied samples 8000 and tolerating samples 1500. Enter the same figures below to reproduce the worked path.
What is Apdex Score?
Apdex turns latency distributions into a 0–1 score using a threshold T you define (satisfied ≤ T, tolerating ≤ 4T, else frustrated).
- Pick T carefully (e.g. 0.5s for APIs, higher for pages)
- 1.00 is perfect; many teams target ≥ 0.85–0.94
- Frustrated samples still sit in the denominator
The Formula
Worked Example
Common Use Cases
- Executive dashboards: one UX-friendly number
- Release gates: minimum Apdex targets
- Endpoint ranking: find painful APIs
Pro Tips
- Publish T with the score
- Do not mix page and API samples
- Track trend, not vanity peaks
Limitations: Apdex Score figures are educational performance-testing aids—not contractual capacity guarantees. Validate with calibrated load tools and production telemetry.
FAQ
Where do frustrated samples go?
They are included in total samples but add 0 to the numerator.
Is Apdex a percentage?
It is a 0–1 index. Multiply by 100 only if your dashboard displays it that way—and label it clearly.
Authoritative References
For performance and quality engineering context, consult:
- ISTQB — performance testing syllabus topics
- W3C Navigation Timing — browser performance timing
- CMG / capacity forums — capacity and performance practice communities