Lena Fischer
Head of Product · Product
LF
ROI score
72Good
Efficiency is flat or down vs last month.
Recommendation
Review what changed vs last month (new model, new task type).
AI spend
$318
5% vs prior period
Yield rate
71%
Team average 61%
Cost per accepted unit
$6.58
Team average $8.83
Leverage vs last month
0.79
68 runs · 48 accepted · 4.4M tokens
Weekly AI spend
Human loops
Accepted85%
Rejected 15% · team 72% accepted
Guidance levelLow
Runs without result16%
Waste$92
ROI score breakdown
Yield 30%
71
Cost efficiency 25%
81
Trend 20%
33
Model fit 10%
100
Capability use 15%
91
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 model23% spend · 8% tokens
Standard model58% spend · 46% tokens
Fast small model19% spend · 46% tokens
Top skills used
- Meeting notes
- Research synthesis
- Translation
Agents and MCP usage
- Docs MCP67 runs
- Web research agent181 runs