What SAP's own customer base is telling itself
ASUG, the SAP user group, surveyed 142 member organizations actively engaging with AI inside their SAP landscapes, sponsored by SAP, Microsoft, and Intel, and the headline finding cuts against a common assumption in enterprise AI discussions. Most respondents are still exploring SAP's AI options rather than running AI enterprise-wide, and the single biggest determinant of whether an organization had moved past pilots was having governance, security, and operational alignment in place, not the sophistication of the AI tools themselves or the cleanliness of their underlying data.
That ordering matters because it inverts the sequence many technology teams assume: get the data right first, then the AI, then worry about governance. ASUG's own researcher, Blake Baltazar, summarized the dominant mindset among the larger group of respondents bluntly: prove it to me before I scale. That is a governance posture and an organizational trust problem well before it is a technology problem.
Two adoption mindsets, and most companies are in the slower one
ASUG identifies two distinct groups among its members. Intentional adopters are strategy-led, deploy AI into live business processes more frequently, and have visible executive involvement driving the effort. Measured adopters, the larger of the two groups, are more experimental by nature and insist on validating AI through pilots before committing to broader deployment, which is the group Baltazar's prove-it-to-me framing describes most directly.
The practical implication for any enterprise benchmarking its own AI maturity against peers is that being a measured adopter is not a sign of lagging behind. It is the more common posture among SAP's own customer base, and the survey's governance finding suggests that caution is, if anything, well calibrated: organizations that validate governance and operational readiness before scaling are the ones more likely to have a smoother path once they do scale.
Where the real friction is, in the group's own words
ASUG CEO Geoff Scott described the broader data readiness challenge across the membership in plain terms: the road has a ton of potholes. That framing matters alongside the survey's finding that data-quality readiness actually ranked lower, at 25%, than governance, security, and operational alignment as a barrier, suggesting that even organizations aware of their data gaps see the organizational and governance work as the harder, more immediate blocker standing between them and enterprise-wide AI.
Scott also flagged a structural shift reshaping how SAP customers operate day to day: gone are the days when we can take parts of SAP and customize them to our heart's content. He argues that public-cloud, multi-tenant S/4HANA is the more sustainable long-term path precisely because heavy ECC customization now carries real support and maintenance costs that compound over time, a tension that sits underneath every AI scaling decision these customers are making on top of their existing SAP landscape.
The resourcing reality underneath the AI conversation
Perhaps the most grounding line in the entire survey write-up is Scott's observation about ordinary workload: most of our member organizations are working 40-plus-hour weeks just getting S/4 done. That is a direct reminder that for a large share of SAP's customer base, AI scaling is competing directly for attention and hours against a core ERP migration that has not finished yet, not against a quiet, fully modernized baseline with spare capacity to experiment.
That resourcing constraint helps explain why only about 10% of respondents had reached the stage of mapping AI's fit to specific teams, roles, and functions ahead of enterprise-wide scaling, with named obstacles including security, privacy, governance, skills gaps, budget constraints, and data readiness all competing for the same limited bandwidth. Enterprises further along in their own S/4HANA migrations should expect to have meaningfully more room to pursue AI scaling than peers still mid-migration, independent of either group's AI ambition.
What this means for any enterprise running SAP, or watching those that do
The clearest actionable signal in the survey is what respondents said would most increase their confidence to scale: 51% pointed to step-by-step implementation support tied specifically to their SAP solutions, ahead of any other factor, including technology improvements. That is a strong argument for enterprise technology leaders to prioritize vendor and partner-delivered implementation playbooks over internally built ones when the option exists, since the user base itself is signaling that generic AI capability is not the bottleneck, specific operational guidance is.
For CIOs outside the SAP ecosystem evaluating their own AI governance maturity, the transferable lesson is to measure readiness by governance and operational alignment first, rather than defaulting to the usual data-quality-first narrative. ASUG's own members, running some of the most data-intensive enterprise systems in existence, still rank governance and operational alignment as the harder problem, which should give any technology leader pause before assuming a data cleanup project is the actual blocker standing between their organization and AI at scale.
