In my previous posts, I shared my personal journey with large language models and how they’ve transformed my productivity. Today, I want to focus specifically on how distributors are using LLMs like ChatGPT and Claude to solve real business problems, based on insights from our recent Distribution Strategy Group workshops and research. 

The Evolution of LLM Capabilities for Distribution 

One of the most exciting developments I’ve seen is how LLMs have evolved to the point where distribution companies now can feed distribution-specific terminology and contexts to allow the tools to be used for real gains in productivity. At our recent AI workshop, we demonstrated how these models can now process and analyze industry-specific information with remarkable accuracy.  

When you ask the right questions, these tools can deliver insights that previously required hours of analysis. For example, when analyzing a strategy, rather than simply asking for general information, experienced users are prompting LLMs with specific questions like: 

“Analyze my product data to identify high-margin items with low current penetration across my customer base, considering seasonality and regional preferences.” 

The level of detail and actionable insight in the responses has improved dramatically from even a year ago.  

From Generic to Distribution-Specific Applications 

Our research shows that LLM adoption has grown substantially, with 63% of distributors now using ChatGPT, along with growing adoption of alternatives like Gemini (5%), CoPilot (3%), and Claude (1%). The frequency of usage is equally telling. Nearly 40% of distribution professionals are using these tools daily, 32% weekly, and 11% monthly, with only 18% not currently using them but exploring options. What’s more interesting than the adoption rates and frequency, however, is how these tools are being used.​​​​​​​​​​​​​​​​ 

Based on our survey data, distributors are applying LLMs in several key areas: 

1. Writing and Content Creation

Product descriptions have long been a challenge for distributors, with many relying on the same manufacturer data that everyone else uses. This creates both a customer experience problem and an SEO disadvantage. 

In our recent workshop, we demonstrated how LLMs can transform generic product information into compelling, unique content. We asked Perplexity about a specific DeWalt tool (DCD793B), it returned comprehensive, ecommerce-ready product descriptions by analyzing information from multiple sources, including the official DeWalt website and major retailers. 

This capability enables distributors to rapidly enhance their product catalog with content that improves both customer experience and search visibility — all without adding headcount to the marketing team. 

2. Process Documentation and Knowledge Management

One of the most practical applications we’ve explored is using LLMs to create and maintain standard operating procedures (SOPs). As demonstrated at our workshop, these tools can: 

• Convert business processes into structured, easy-to-follow documentation 

• Organize FAQs and reference materials into searchable knowledge bases 

• Develop comprehensive onboarding programs for new hires 

• Create customized training materials for different roles 

This is particularly valuable for distributors facing workforce challenges and knowledge transfer issues as experienced staff retire. This is a way to use LLMs to document “tribal knowledge” from retiring veterans, creating a permanent repository of product expertise and customer insights. 

3. Sales Support and Customer Service

The most impactful application is in sales and customer service. During my demonstration, I showed how LLMs can analyze customer communications and automatically draft responses to common scenarios, allowing representatives to focus on relationship building rather than routine communications. These tools not only allow for more personalized interactions but make it easier for the service teams to respond to customers. 

For example, when presented with a scenario involving incorrect deliveries during a promotion, an LLM generated a comprehensive analysis of potential causes and drafted an empathetic response that acknowledged the issue while outlining concrete steps to resolve it. All within seconds. 

The tool also suggested targeted questions to better understand the root cause, ensuring the response wasn’t just sympathetic but actionable. This kind of assistance enables even newer team members to respond with the expertise of seasoned professionals. 

4. Market Research and Competitive Analysis

Another powerful application is using LLMs for market research and competitive intelligence. Every distribution executive tries to keep up with what’s happening in the market, but with the plethora of information that’s out there, it’s an impossible task to do it alone. 

At our workshop, I demonstrated how these tools can analyze industry trends, competitive positioning and customer preferences to support strategic decision-making. 

When asked to analyze the convenience store industry, an LLM generated a structured report covering key market trends, competitor analysis and customer insights — organized in a clear, accessible format. This type of analysis, which might previously have required days of research, was completed in minutes. 

Focus on Prompt Engineering 

The quality of output from any LLM depends heavily on the quality of input. As we demonstrated in our workshop, effective prompt engineering is a crucial skill for getting the most value from these tools. This is a skill that must be effectively trained throughout your organization. 

The most effective prompts include four key elements: 

• Clearly defined roles (who you want the LLM to be) 

• Specific goals using action words 

• Relevant context and background information 

• Desired output format specifications 

At DSG we created libraries of effective prompts for common business scenarios, allowing team members to leverage proven approaches rather than starting from scratch. 

Creating an LLM Implementation Strategy 

Based on the experiences shared at our events, I’ve developed a framework for effective LLM implementation in distribution: 

• Start with specific business problems rather than technology infatuation. Focus on areas like product descriptions, quote generation or customer service where repetitive knowledge work creates bottlenecks.  

• Develop a contained pilot project with clear success metrics where subject matter experts can validate outputs.  

• Invest in prompt engineering expertise by creating templates and libraries of effective prompts for common scenarios.  

• Integrate capabilities directly into existing workflows rather than creating separate systems.  

• Ensure leadership actively uses and champions these tools. Executives who personally engage with LLMs drive significantly faster adoption and better results across their organizations. As one distribution leader noted in our panel: “The most powerful implementation strategy isn’t about the technology – it’s about using these tools to solve real problems that impact the bottom line.”​​​​​​​​​​​​​​​​ 

Establish Governance and Best Practices 

As with any technology, establishing clear guidelines for appropriate use is essential. Our research shows that 13% of distributors are educating themselves on IT security needs related to AI, with another 47% making some effort in this direction. 

Successful organizations are developing policies that address concerns such as data privacy, appropriate content and quality control while still encouraging innovation and experimentation. 

The Competitive Imperative 

The adoption curve for LLMs in distribution is rapidly accelerating. Our survey shows that 33% of distributors expect a significant increase in generative AI usage over the next two years, with another 35% anticipating moderate growth. Only 7% expect no increase. 

This trajectory means that distributors who delay implementation risk falling behind competitors who are already capturing efficiency gains and enhancing customer experiences.  

Learning from Early Adopters 

Distributors seeing the greatest success with LLMs share several common characteristics: 

Executive Engagement: Leadership teams that actively use and champion these tools drive faster adoption and better results. If you are a leader and you’re not using these large language models, then you’re doing your organizational disservice 

Structured Experimentation: They encourage controlled experimentation with clear success metrics. Your team is using these models, whether you believe it or not so engaging these individuals in a structured way helps you get better at these models faster. 

Continuous Learning: They invest in ongoing training and knowledge sharing about effective LLM utilization. The models are changing so rapidly you need to be able to keep the pace of change. Having dedicated team to help will ensure you are using the latest models for your organization. 

My Journey Continues 

As I continue exploring and implementing these technologies, I’m more convinced than ever of their transformative potential. That 35% productivity gain I mentioned in my first post isn’t theoretical — it’s my actual experience, and it’s changed how I approach my work. I will begin to share specifics around each of the individual large language models in the coming updates. 

I’m looking forward to continuing this conversation at our Applied AI for Distributors conference in June, where we’ll be showcasing practical applications of AI technologies specifically designed for distribution operations. 

Join me and other distribution leaders to explore practical LLM applications at the Applied AI for Distributors conference, June 24-26 in Chicago. Learn more.