Two surprising paradoxes in sales AI adoption

As I reflect on AI adoption across sales and revenue organizations over the past year, two paradoxes stand out.

First, data platforms are taking a back seat. While data is the undisputed foundation of any AI deployment, most sales organizations show limited concern for the data platform they are actually putting together. I expected data platforms to move to center stage in 2025 and become a key influencer of technology decisions. That has not happened. Sales and operations teams are instead prioritizing specific AI use cases they can deploy, with little scrutiny of the embedded data platform bundled with the chosen solution. Only the most data-mature organizations push vendors to connect to their data warehouse, and even then, the discussion rarely goes further.

Second, while organizations consistently complain about solution stack complexity and tool sprawl, many AI deployments, particularly agents, focus on stitching together existing stacks. Tool consolidation has been a long-standing aspiration for several years. Until last year, it was constrained by the limited maturity of integrated solutions. Fast forward to 2025, revenue platforms such as Apollo, Clari-Salesloft, Gong, Outreach, and ZoomInfo, followed by new entrants and CRM providers, have not only significantly matured but have also leveraged AI to accelerate their product roadmaps even further.

I was surprised to see this “overlay” approach gaining such traction. It is somewhat understandable, though: most organizations are targeting specific use cases for their initial AI deployments, and agents excel at simple workflow automation. The overlay approach delivers a double win. It allows sales organizations to finally do things they long wanted, such as automating CRM data entry or surfacing priority leads to engage with, while also maximizing their existing technology investments.

Another factor explaining these paradoxes is that the SalesTech industry remains overindexed on serving SaaS and technology companies. This segment is dominated by startups and scale-ups that prioritize speed and action over well-designed technology stacks.

I have come to terms with these two paradoxes, which are likely to persist into 2026. The irony is that recognition is growing around the need for a middle layer between data and AI agents to unify datasets, with context graphs emerging as the next frontier for AI effectiveness. The pendulum will likely swing back over time, refocusing attention on building a unified data foundation and simplifying sales stacks around one of the emerging revenue suites.