Distribution’s AI Moment Has Arrived. Are You Ready?
A Conversation with David Wascom, SAP Industry Executive Advisor for Wholesale Distribution
Every so often, an industry reaches a moment where the cost of standing still exceeds the cost of moving forward. For wholesale distribution, a business model largely unchanged for 300 years, that moment is now.
I sat down with David Wascom, SAP Industry Executive Advisor for Wholesale and Distribution, for a two-part conversation that ranged from the history of the industry to the mechanics of AI governance to a pointed call to action. What follows is a synthesis of the most important insights from that discussion, including direct quotes from David.
An Industry Built on a 300-Year-Old Model
Distribution is, at its core, a simple business. Buy in bulk, break it down, sell to business customers. David pointed out that this model has been largely unchanged for roughly three centuries.
“If you took somebody from the mid-1800s and brought them forward into a late 1990s distributor, other than the data entry portion, they’d feel pretty comfortable with the model.”
But the last 20 years have introduced significant change. Suppliers are demanding more from distributors. Customers expect more than inventory. They want solutions, anticipation, and platforms. And the macroeconomic environment (supply chain disruptions, pandemics, tariffs, political volatility) has made the stakes of operational efficiency higher than ever.
The industry has been slow to adopt new technology, and David is direct about why: when your model has worked for 300 years, there is no urgency to change. But urgency has arrived.
What the Autonomous Enterprise Means for a Distributor
SAP Sapphire 2026 was headlined by the “autonomous enterprise,” a vision of AI that goes well beyond chatbots and information retrieval. David broke it down into three distinct waves:
- Informational AI: Ask a question, get an answer. Think early ChatGPT use cases.
- Agentic AI: Give an agent a defined task. Process these payments. Match these invoices. Fast, rules-based execution.
- Orchestrated AI: Multiple agents coordinating across an entire business process, operating with broader authority and judgment, with a human in an oversight role rather than a hands-on execution role.
For a distributor, this translates to a fundamental shift in what people do. Not fewer people, but people focused on higher-value work: monitoring system behavior, refining parameters, identifying new opportunities, and advancing the business.
“The traditional distributor response has always been to throw more people at it. Distribution has got to be about optimization, about efficiency.”
Three Places to Start Right Now
David identified three areas where AI can deliver near-term, measurable value for distributors, and where solutions exist today:
1. Sales Order and Quote Automation
In trade distribution (electrical, plumbing, HVAC, building products) first response often wins. Getting a 100-line-item quote turned around faster than a competitor is a genuine competitive advantage. AI agents can process unstructured inbound orders (David’s example: a photo of handwritten notes on a napkin), match part numbers, pull pricing, and return a formatted quote in a fraction of the time a human would take.
DataXstream’s OMS+ includes exactly this capability. David specifically called it out as a strong example of this kind of solution in action.
2. Finance and Accounts Receivable
Processing high volumes of customer payments, identifying which accounts need human attention, and building dunning plans are all tasks that AI handles well. The value is not eliminating your AR team. It is letting agents process the straightforward 970 accounts so your people can focus on the 30 that need a real conversation.
Inventory optimization, sourcing decisions, warehouse management, fleet scheduling. These are algorithm-native problems. The challenge historically has been the cost and complexity of implementing them. That barrier is dropping fast.
Governance: Managing Agents Like You Manage People
One of the most thought-provoking parts of our conversation centered on governance: the idea that AI needs to be explainable, auditable, and bounded.
David used a memorable analogy: he recalled watching buyers walk the warehouse floor with clipboards, making procurement decisions based on whether the shelves were empty. Effective? Possibly. And yet replacing that entirely with an unchecked system creates its own category of disaster. He argues for keeping a human in the loop, to manage the AI agents and make adjustments if results are not desirable.
The governance conversation is also about trust. If an agent returns an answer without an explanation to how it got there, it may be difficult to trust the result or to repeat it. SAP’s push for explainability is not a compliance checkbox. It is foundational to making AI useful with repeatable processes that return reliable answers.
Connected Data: Context Is Everything
A clean data environment is a nice-to-have. A contextualized one is a requirement. David made this distinction clearly: it is not about whether every zip code in your database is correct. It is about whether your organization understands where the data comes from, what it represents, and how it relates to everything else.
He used a sharp analogy: a newscaster reading sports scores (“3 to 1, 14 to 7, 7 to 0”) without identifying the sport or the teams. Numbers without context are noise.
This is why SAP has been pushing the Business Data Cloud concept: not just data storage, but the business context that makes data actionable across procurement, sales, warehousing, and finance simultaneously.
The Haves and the Have-Nots
David cited research from Distribution Strategy Group: more than 80% of distributors are experimenting with AI, but only a single-digit percentage are delivering real business value from it. The rest are “splashing,” as David put it, making noise without going anywhere.
The distributors who are moving, building governance frameworks, documenting their processes, educating their people, and piloting real use cases, are not necessarily the largest or best-resourced companies. This is not a “big guys only” technology. The entry points are accessible. The solutions exist. The question is whether you are willing to get off the fence.
“You don’t have to start with a blank whiteboard. A lot of the work has already been done. The trick is to understand what’s going to be important from your business standpoint.”
What You Can Do Right Now, Regardless of Where You Are
David’s practical advice, distilled:
- Build a business strategy that uses AI, not an AI strategy. What problems are you trying to solve? Start there.
- Build a data governance framework. Understand who owns what data, where it comes from, and what it means.
- Document your processes. You cannot improve what you have not defined. Documented processes also become input data for AI agents.
- Educate your people. Your team is already using AI. Help them use it well, within a governed framework.
- Start a pilot. Pick one of the three entry points above. Find a partner. Move.
Watch the Full Conversation
This post is drawn from a two-part video interview with David Wascom recorded in June 2026. If you want to hear it directly from David, including the clipboard-and-warehouse analogy and the math teacher moment, the full video is worth your time.
