Why micro-TLX works
Full NASA TLX has 6 dimensions and takes 3-5 minutes. Teams skip it. Micro-TLX captures mental demand + frustration in 15 seconds. Teams actually do it. Compliance beats precision.
The Two Sliders
Slider 1: Mental Demand (0-100)
Question: "How mentally demanding was this task?"
| Range | Meaning | Example | |-------|---------|---------| | 0-20 | Almost automatic | Routine email, copy-paste task | | 20-40 | Light effort | Familiar code review, standard doc | | 40-60 | Moderate focus | New problem, some complexity | | 60-80 | Heavy concentration | Complex debugging, strategic analysis | | 80-100 | Exhausting | Crisis response, unfamiliar domain |
Slider 2: Frustration (0-100)
Question: "How frustrated did you feel?"
| Range | Meaning | Example | |-------|---------|---------| | 0-20 | Calm, flow state | Task went smoothly | | 20-40 | Minor annoyances | Small blockers, easy fixes | | 40-60 | Noticeable friction | Tool issues, unclear requirements | | 60-80 | Significant stress | Repeated failures, time pressure | | 80-100 | Very frustrated | Blocked, overwhelmed, angry |
When to Ask
- After every AI-assisted task — before you move to the next thing
- Before breaks — capture state while it's fresh
- When switching contexts — especially high-stakes to routine
- End of focus blocks — 90-minute Pomodoro or equivalent
What the Numbers Mean
Green Zone (both under 40)
You're in flow. Current pace is sustainable. Keep going.
Yellow Zone (one or both 40-60)
Fatigue is building. Consider:
- Take a 5-minute break before next task
- Switch to easier work
- Check if AI is helping or adding friction
Orange Zone (one or both 60-80)
Performance is likely degrading. Actions:
- Stop after current task
- Take a real break (15+ minutes)
- Review: Is AI making this harder?
- Consider delegation or delay
Red Zone (either over 80)
Stop immediately. You're making mistakes you can't see. Actions:
- Step away from the task
- Don't make important decisions
- Debrief: What caused this spike?
- Reset before returning
- ✓Log micro-TLX after every AI-assisted task
- ✓React immediately to yellow/orange/red zones
- ✓Track patterns: Which tasks spike TLX?
- ✓Share extreme readings with your team lead
- ✓Pair TLX with Δ to see full picture
Reacting to Patterns
High mental demand + low frustration
The task is hard but you're coping. This is good challenge. Monitor for accumulation across the day.
Low mental demand + high frustration
Tool problems, unclear requirements, or friction with AI. Fix the process, not yourself.
Both high
Danger zone. The task is hard AND the process is broken. Escalate or delegate.
Both low
Either the task is too easy (could you automate more?) or you're underestimating load (check quality of output).
“"We caught three burnout situations early because engineers logged TLX honestly. The pattern was obvious in the weekly chart."”
Related Resources
- TLX Help Guide — deeper dive into TLX methodology
- How to Read Your Results — interpret TLX alongside Δ
Apply this now
Practice prompt
Log micro-TLX after your next three AI-assisted tasks and note the pattern.
Try this now
Run any Analyzer pack and pay attention to the TLX prompts—they use this framework.
Common pitfall
Logging TLX but not reacting to it—the value is in the immediate response.
Key takeaways
- •Two sliders, 15 seconds—mental demand and frustration only
- •React immediately: green = go, yellow = slow, orange = pause, red = stop
- •Track patterns to identify which tasks or tools cause spikes
See it in action
Drop this into a measured run—demo it, then tie it back to your methodology.
See also
Pair this play with related resources, methodology notes, or quickstarts.
Further reading
Next Steps
Ready to measure your AI impact? Start with a quick demo to see your Overestimation Δ and cognitive load metrics.