SAP makes a double play in data and AI with acquisitions of Prior Labs and Dremio

SAP announced two acquisitions, Prior Labs and Dremio, in a move backed by over one billion euros of investment to strengthen its AI stack for structured business data. The deals, expected to close in 2026 subject to regulatory approval, should enhance SAP Business Data Cloud and SAP AI Core with tabular foundation models and an Apache Iceberg-native data layer. SAP is also preserving both firms' open-source strategies, while building toward a universal catalog and Knowledge Graph for enterprise AI.

Analysis

This is less about “AI spend” and more about SAP trying to re-anchor enterprise software around proprietary workflow intelligence before horizontal model vendors commoditize the interface layer. The strategic edge is that SAP controls both the transactional substrate and the user workflow, so if it can make structured-data models good enough, it can tax the entire inference stack via switching costs rather than raw model performance. That is a meaningful moat expansion, but it also raises the bar: the market will now expect measurable monetization inside cloud ARR, not just narrative lift.

The biggest second-order winner is not necessarily SAP’s own AI revenue, but the surrounding ecosystem of data tooling and governance vendors that plug into open formats and catalogs. If SAP successfully makes Iceberg-style interoperability the default, the losers are proprietary data-platform incumbents whose lock-in economics depend on format friction and bespoke pipelines. The real competitive risk is that hyperscalers and model labs will respond by bundling similarly “good enough” structured-data agents into broader cloud contracts, pressuring SAP to prove it can charge a premium for vertical specificity rather than becoming an orchestration layer with modest take-rate.

Catalyst timing matters: the stock can rerate on announcement, but the harder proof point is 2H26 when integration, regulatory closure, and actual product adoption intersect. The near-term upside is sentiment-driven; the medium-term downside is execution risk if the acquisitions remain independent assets that are difficult to unify into a coherent developer experience. If SAP fails to show conversion from AI features into higher net retention or larger deal sizes over the next 2-3 quarters, the market will likely reclassify this as expensive capability expansion rather than accretive platform leverage.

Consensus is probably underestimating how much this raises the strategic value of SAP’s installed base in a world where general-purpose AI becomes cheap. The more interesting contrarian take is that this is defensive as much as offensive: SAP is trying to prevent customers from building a thinner AI layer on top of SAP data without paying SAP for the privilege. If successful, that supports a higher long-duration multiple; if not, it simply funds open-source ecosystem growth and leaves SAP with more capex and more integration complexity.

Key Decisions for Investors

• Add SAP on 3-6 month dips, with a 12-month horizon, targeting a re-rating if management begins quantifying AI-driven cloud attach; risk/reward improves if the stock fades post-announcement despite no change in core demand.
• Buy SAP call spreads expiring after the expected 2H26 integration milestones (6-12 month tenor) to express upside from successful execution while capping premium at risk; this is preferable to outright equity if you expect volatility around regulatory closure.
• Pair long SAP / short a horizontal enterprise AI beneficiary with weaker vertical distribution, to isolate the value of proprietary workflow access versus generic model exposure; this works best if enterprise AI spend rotates from experimentation to procurement.
• Avoid chasing pure-play data-platform names on the open-architecture theme until SAP demonstrates actual adoption metrics; the trade likely works only if capital flows to ecosystem enablers faster than incumbents can defend share.
• If SAP rallies sharply on headlines, consider trimming into strength and waiting for evidence of ARR or margin contribution; the main risk is multiple expansion ahead of fundamentals, which can retrace if integration headlines slow.