For most of the past two decades, enterprise software value followed a predictable pattern. Vendors built platforms, customers bought licenses, and Partners implemented the solution. Each wave of technology changed the tooling but not the structure of responsibility.
AI breaks that pattern.
Unlike previous platform transitions, organizations are discovering that adopting the technology does not automatically produce usable outcomes. Systems can be deployed successfully yet fail operationally. The limiting factor is no longer access to software but the ability to adapt workflows around it.
At that point in the conversation, implementation stops being technical and becomes interpretive. That is where the Partner role changes.
Cecilia Flombaum, Microsoft’s Ecosystem Lead for Business & Industry Solutions, works with Partners on large-scale AI deployments across industries. From that vantage point, she says she sees the shift less as a new product cycle and more as a structural change in how customers operate.
AI is moving very fast
Enterprise technology historically evolved in multi-year adoption cycles. AI is compressing those cycles into months.
The McKinsey State of AI report finds 88 percent of organizations already use AI in at least one business function, and most expect to increase investment in the next three years. A separate McKinsey analysis of AI data-center growth projects roughly 33 percent annual expansion through 2030.
Flombaum said many organizations are still reacting as if the industry has years to adjust. Partners, she said, have not fully absorbed “the velocity of this transformation.”
Even relative to cloud transformation, the acceleration is visible. Investopedia’s coverage of generative AI adoption trends reports AI businesses scaling faster than early cloud platforms at similar stages.
Organizations are therefore adopting technology faster than they can operationalize it. The gap that creates is organizational rather than technical.
The AI transformation is bigger than the move from on-prem to cloud
The cloud migration changed where software ran. AI, however, changes how work itself happens.
Flombaum described the distinction as customer-facing rather than infrastructural. AI, she said, is “revolutionizing how users and software interact.” In contrast, the on-prem-to-cloud shift primarily affected where data was managed.
A study on AI decision-making in organizations shows companies struggle less with deploying models than with adapting processes around them. Likewise, a Gartner forecast on AI-ready data finds many initiatives fail because organizations are not prepared to operate AI systems.
For Partners, that shifts the work toward guided change. Flombaum said customers now need help learning “a different way of working.”
Customers are experts in their field, not tech
Most organizations understand their industries. They rarely understand how probabilistic systems should reshape workflows inside those industries.
An analysis of AI project outcomes identifies unclear objectives and a misunderstanding of AI capabilities as primary causes of failure. Companies know what they want to improve, but not how AI should change behaviour to achieve it. Flombaum said Partners must be the ones to bridge that gap.
Beyond implementation, Partners must develop business transformation skills to guide customers through operational change.
Customers still provide domain expertise. Partners increasingly translate it into system behaviour.
Modern tech requires data scientists
In agent implementations, Flombaum said organizations routinely underestimate the effort required to reach reliable performance. Building and evaluating custom agents requires specialized expertise and “is so much harder than people think.”
Industry research reaches the same conclusion. A McKinsey report on data-driven enterprises shows value emerges only when models, workflows, and curated data operate together rather than when software is simply deployed.
The result is a blended role across implementations. Flombaum says Microsoft has already adopted a profile combining developer, consultant, and data scientist responsibilities within a single engagement team.
AI is faster, more operationally disruptive, and more dependent on interdisciplinary expertise than previous platform shifts. Customers understand their business problems. Vendors build platforms. But neither alone reliably produces outcomes.
The Partner, instead, becomes the mechanism that makes AI usable.
