Thoughts on AI in chemical distribution: what's working, what to skip, and how operators are putting it to use.
The suppliers with the best brand recognition are often the ones giving you the worst margin because they already sold the customer before you did. I pull gross margin by line to see which suppliers actually need me, and which ones are just counting my money.
My dad arrived in the United States from Sicily as a little boy, 74 years ago today. A look back at the bet my grandparents made on a country they had never seen, and everything my family built because of it.
Surfactants may be the hardest product category for a chemical distribution rep to sell, with HLB values and cloud point behavior that don't fit into a quick pitch. Distributors building searchable, structured product knowledge for their reps are turning that complexity into a real advantage.
Inflation pushed distributor prices up and covered for what wasn't working underneath, manual order entry, bad ERP data, non-strategic SKUs. When the pricing cycle turns, and it always does, those problems show up on the P&L again, so the winners are using today's profits to fix operations before margins compress.
Distributors keep hunting for a mythical sales hire who knows chemistry, can close, and loves to travel, while the bench chemist and QC tech already on staff speak that language with customers. Chemistry takes years to learn; sales and CRM can be taught, so hand them a territory instead of posting another job.
Chemical distribution has the lowest AI exposure of any sector, just 14%, even though the tools already work. Every company I talk to is stuck in the same standoff, the CEO waiting on the team, the team waiting on permission, and nobody actually owning the first project.
I compared GPT-5.5, Gemini 3.1 Pro, and Claude Opus 4.7 on what each is actually good at, ChatGPT is the most complete out of the box, Gemini wins on ecosystem and multimodal work, and Claude wins on writing, reasoning, and coding. Pick the tool based on what the work actually requires, not brand loyalty.
Anthropic raised $65 billion in one week, on top of Google's $40 billion and Amazon's $25 billion announced the same week. Distributors with clean data and a clear problem to solve will move faster than the ones waiting to make a move.
After 25 years in chemical distribution, I've seen this fear before. New technology frees people up to work at a higher level, and the reps using AI to prep for calls are becoming the most valuable people in the room.
Chemical distributors want to adopt AI, but their data is scattered across the ERP, Outlook, a spreadsheet, and a shared drive nobody's opened in a year. Organizing that data is the real first step, before picking any AI tool.
Proper call prep takes 45 minutes and nobody has time for eight of those a day, so I built a Claude agent that checks my calendar, pulls HubSpot history, and emails a full precall brief by 7am. When prep takes zero effort, every rep actually prepares.
Executives spend their day in email, not meetings or strategy, and Outlook can't keep up with that pace. I switched to Superhuman for keyboard shortcuts, AI drafted replies in my voice, and a split inbox that separates signal from noise.
