Every SAP customer is asking where AI will deliver value. Most answers start from the technology. This approach starts from your own system: what it runs, how each process performs, and where an AI agent is actually the right tool.
A catalog of agents or a workshop wish list is hard to scope and harder to defend to a CFO.
Without process KPIs, nobody can say which gap is worth closing first, or prove it was closed.
Many of the most valuable fixes are a rule, a data cleanup or a calculation.
The model uses stated rules to decide. They are visible to the customer, not buried in a black box.
Scroll the table sideways, or open the full-size image.
Sample output for an illustrative ECC customer. A real roadmap is generated from the customer's own measured KPIs.
Each item shows the approach, the KPI gap behind it, and which value levers it moves: cost, revenue, working capital, and risk & control. Items are grouped into waves: Now, Next, Later, plus an automation track.
Of the 32 gaps, 10 are strong agent candidates, 14 are hybrids, and 8 should stay deterministic.
Every process gap maps to the APQC Process Classification Framework and, where relevant, the SCOR Digital Standard. KPIs use APQC Open Standards Benchmarking measures and SCOR metrics, with SMRP, DAMA-DMBOK, COSO and ITIL where those apply. Benchmark values are licensed by their publishers and are not reproduced; customers enter their own targets.