BARC x Actian Research: Why Data Products are Key to Scaling AI
Résumé
- Organizations with mature data products are far more likely to scale AI, with 85% running multiple AI projects in production.
- Agentic AI adoption is significantly higher in these organizations, with 77% deploying autonomous systems versus limited adoption elsewhere.
- Data products provide structured, governed, and reusable data foundations required for reliable production AI.
- Strong data ownership, quality standards, and contracts enable safer and more scalable AI operations.
Conclusions de l'étude mondiale BARC x Actian de 2026, menée auprès de plus de 300 responsables des données en entreprise
Dans tous les secteurs, les entreprises investissent massivement dans l'IA. Pourtant, de nombreuses initiatives peinent encore à dépasser le stade de l'expérimentation et à générer une valeur ajoutée durable pour l'entreprise.
Pour mieux comprendre ce qui distingue les initiatives d'IA qui se développent de celles qui stagnent, Actian s'est associé à BARC, un cabinet d'analyse mondial de premier plan spécialisé dans les données et l'analyse, pour mener une étude mondiale auprès des responsables des données d'entreprise.
S'appuyant sur les témoignages de plus de 300 personnes interrogées issues de divers secteurs et régions, cette étude examine la manière dont les organisations adoptent et mettent en œuvre data products data contracts, ainsi que l'influence de ces approches sur la réussite des initiatives en matière d'IA.
One of the clearest insights emerging from the research is the strong relationship between data product adoption and the ability to scale AI into production.
Rapport mondial de recherche BARC × Actian (2026)
Organizations With Mature Data Products Run More AI in Production
Many organizations today are experimenting with AI. However, moving AI initiatives from experimentation into production remains difficult.
The research reveals a clear pattern: organizations that deploy data products at scale run significantly more AI systems in production.
Among organizations that deploy data products company-wide, 85% report running multiple AI projects in production. By contrast, organizations that are not using—or only experimenting with—data products are far less likely to operationalize AI beyond pilot initiatives.
This finding highlights the role data products play in enabling reliable AI systems. By organizing data assets with clear ownership, defined quality expectations, and reusable interfaces, organizations create the structured data foundation required to support production AI.
Correlation between AI project adoption and data product usage – © BARC 2026
Agentic AI Adoption is Highest in Organizations With Mature Data Products
The same pattern appears when looking at agentic or autonomous AI systems.
Agentic AI systems—capable of acting autonomously, orchestrating workflows, and interacting with multiple data sources—require highly reliable and well-governed data environments to operate safely.
The research shows that 77% of organizations with company-wide data products already have at least one agentic or autonomous AI system in limited or full production. The inverse also holds true: 77% of organizations that are only experimenting with data products—or not using them at all—have not yet deployed agentic AI in production.
This finding reinforces an important insight: as AI systems become more autonomous and integrated into operational workflows, the quality, governance, and reliability of underlying data become even more critical. Data products provide the structure needed to manage these dependencies and support the safe deployment of increasingly complex AI capabilities.
Correlation between agentic AI adoption and data product usage – © BARC 2026
Why Data Products Enable AI at Scale
The correlation observed in the research reflects the role that data products play in modern data architectures.
By organizing data assets as products—with defined ownership, quality standards, and clear interfaces—often reinforced through data contracts—organizations create a reliable foundation for analytics and AI systems.
This approach helps ensure that AI systems rely on trusted, well-managed data, reducing the risks associated with inconsistent inputs, unclear responsibilities, or fragile data pipelines.
As a result, data products are emerging as a critical capability for organizations seeking to operationalize AI at scale.
Télécharger le rapport de recherche complet
Cet article présente plusieurs conclusions clés issues de l'étude mondiale menée par BARC et Actian.
Le rapport complet, Data Products Data Contracts 2026 : les fondements du succès de l'IA, examine comment les organisations de tous les secteurs et de toutes les régions adoptent et mettent en œuvre data products data contracts, et comment ces pratiques influencent la maturité de l'IA, gouvernance et les résultats concrets.
Téléchargez le rapport complet pour découvrir les conclusions de cette étude et savoir comment les grandes entreprises déploient l'IA à grande échelle grâce à data products.
Télécharger le rapport complet
Restez connecté
Des informations exploitables à votre disposition.
(par exemple, sales@..., support)
