Surveys Reveal Gap Between AI Ambition and Operational Reality in Distribution

Over the past year, ASA has gathered input from members through two Applied AI surveys, one in October 2025 and another in April 2026. The goal was not to measure adoption precisely, but to understand how companies in distribution are actually working through AI and where it is showing up in real operations.
 

The signal is clear, and it shows up consistently across both surveys. Companies are interested, and many are actively exploring or piloting AI. At the same time, the way it is being used today does not always line up with where companies expect it to have the biggest impact.
 

When members describe where they want to apply AI, the focus is heavily on revenue-facing areas. Customer intelligence, forecasting, and sales enablement come up most often, and those are the places where companies expect better decisions, stronger customer relationships, and growth.
 

But when members describe the challenges they are dealing with day to day, a different pattern shows up. Manual and repetitive work is the most common issue, followed closely by data quality and integration challenges. Customer insight, reporting, product data, and system alignment all come up as areas where work slows down or requires additional effort.
 

The gap between those two views is where most of the friction sits. Companies are not struggling to imagine where AI could be useful, but they are running into the reality of how their data and systems are structured when they try to apply those ideas in a practical way.
 

This shows up quickly once teams begin working with real use cases. A group might start with the goal of improving customer visibility or forecasting demand, but as they begin working with the data, inconsistencies appear. Customer records do not align across systems, product data is incomplete or structured differently depending on the source, and information that should connect does not always do so cleanly.
 

At that point, the focus shifts. Instead of analyzing trends, the team spends time reconciling data, and instead of identifying opportunities, they work to make the information usable. The tools are still helpful, but they are being used to address underlying issues rather than to move directly into higher-value use cases.
 

This pattern is not limited to sales. Inside sales teams see it when quoting requires additional verification of product information, finance teams encounter it when trying to automate reconciliation across systems that do not align cleanly, and operations runs into it when inventory, purchasing, and product data do not support more advanced planning without manual intervention.
 

Across the business, the same thing happens. AI is introduced with the expectation that it will drive new insight or efficiency, and it ends up being pulled into the work that already slows the organization down.
 

That does not mean progress is not happening. In many cases, this is exactly what progress looks like, even if it does not match the original expectation.
 

The surveys show that companies are moving beyond awareness and into real application. They are testing, learning, and beginning to connect AI to actual workflows, but the first impact is often in making existing processes more manageable rather than transforming them immediately.
 

This is where expectations need to adjust. The most visible use cases tend to be tied to growth and performance, while the work that enables those outcomes is less visible but necessary for those use cases to function at all.
 

Companies that are moving forward are not skipping this step. They are starting where the friction already exists and using AI to reduce the effort required to manage it, whether that means improving how product data is structured, supporting teams with repetitive tasks, or making information easier to access and interpret.
 

For leaders, this changes where to focus. Instead of starting with a broad question of what AI could do, it becomes more practical to look at where the business is already slowing down and where teams are spending time fixing data, interpreting information, or repeating the same steps.
 

From there, progress becomes easier to build. A specific use case can be identified, tested, and refined, and the data supporting it improves as part of that process, allowing those improvements to carry forward into other areas of the business over time.
 

This is the transition the industry is working through right now. ASA’s Applied AI work is focused on supporting that shift by reflecting what is actually happening across distribution and helping translate it into practical starting points.
 

Members have been clear that they want to see what works in real environments, and that is where the focus is going. AI is not short on potential in distribution, but the path forward is being shaped by how companies work through the challenges that show up first.

By Nils Swenson