Hi,
Field Note #4. Today's case: an AI decision-support layer for a DACH mid-market enterprise running a 40-initiative transformation portfolio out of Excel and monthly steering committees.
We pitched "replace LeanIX." Got nowhere. Their data steward already kept LeanIX clean. Then we pitched "sit on top of it" — read initiatives, score them, write back as properties. That worked.
What's inside:
• Initiatives scored before steering committee: 30% → 100% in Q1
• Time-from-list-to-ranked-priorities: 2-3 weeks → 2 days
• Steering committee meetings spent on prioritization: 2.5/month → 0.5/month
• The floor rule that caught two would-be greenlit initiatives (no Greenlight if any dimension < 3)
• Eval AUC 0.83 against the CIO's historical rankings — that's what bought trust
• Full architecture + 8-step checklist to layer this on your EAM tool