Felix Wagner
Support Specialist · Customer Support
FW
ROI score
62Watch
Efficiency is flat or down vs last month.
Recommendation
Review what changed vs last month (new model, new task type).
AI spend
€413
9% vs prior period
Yield rate
65%
Team average 61%
Cost per accepted unit
€3.58
Team average €3.44
Leverage vs last month
0.95
177 runs · 115 accepted · 7M tokens
Weekly AI spend
Human loops
Accepted72%
Rejected 28% · team 73% accepted
Guidance levelMedium
Runs without result10%
Waste€232
ROI score breakdown
Yield 30%
65
Cost efficiency 25%
63
Trend 20%
46
Model fit 10%
67
Capability use 15%
70
Each component is scored 0–100; 50 means on par with the team median or flat vs last month. Good ≥ 70, Watch 40–69, Bad < 40.
Model mix: spend share vs token share
Premium reasoning model52% spend · 18% tokens
Standard model36% spend · 29% tokens
Fast small model12% spend · 53% tokens
Top skills used
- Spreadsheet automation
- Image generation
- Summarisation
Agents and MCP usage
- Docs MCP115 runs
- Web research agent22 runs