Discover the common pitfalls of AI in the chemicals industry and learn how to align initiatives with business goals for successful outcomes.
... and One Way to Win -
A CEO comes back from an industry conference energized. Within a month, there are six new working groups. Each function has identified its own priorities. There is no unifying logic connecting any of it to strategic outcomes, no financial objective for any individual initiative, and no one in the business who is accountable for results. The CIO is coordinating the portfolio. Progress is measured in pilots launched, not value delivered.
This pattern isn't rare. It's the default, and in a prolonged trough it's expensive in ways that compound before they're visible.
The failure modes for AI programs in chemicals are predictable enough that I can describe them before they finish playing out. There are three, and they rarely appear alone.
The Three Failure Modes
The first is pilot proliferation without a financial objective. The executive team is bought in, the working groups are active, the proof-of-concepts are running, but no individual initiative is tied to a P&L outcome. Nobody can answer the basic question: how much is this worth, and when does it show up? When budget conversations get uncomfortable, and in this market they will, these initiatives can't defend themselves. They get cut not because they failed, but because no one can say whether they succeeded.
The second is activity mistaken for progress. Use-case inventories get built. Pilots complete. Steering committee decks get presented. But the work is happening one layer removed from the business, evaluated on technical criteria rather than commercial ones. The measure of success becomes "did we learn something?" instead of "did it move the P&L?" The learning is real. The P&L impact is not.
The third is handing the agenda to technology. The AI program gets assigned to the CIO. That's not a failure of the CIO; it's a structural problem. When technology owns the roadmap, investment decisions get made toward what's technically feasible rather than what's strategically valuable. Business functions become requestors rather than owners. The commercial and operational leaders who need to change how their work gets done are not in the room.
None of these failure modes requires bad intentions or weak leadership. They are the natural output of treating AI as a technology deployment.
Actioning the Insight
What actually works is a different pattern entirely.
The organizations capturing AI value at scale have one thing in common: the initiative is business-led. The roadmap connects AI investments to specific strategic priorities and P&L outcomes. A small number of high-impact domains get real focus, rather than a dispersed portfolio of exploratory efforts. Business and technology teams work together from the start, not in sequence. Talent is built by developing people who already understand the business in context. And adoption is managed actively: incentives aligned, behaviors tracked, outcomes measured.
Most ambitious companies are still following the failure pattern. The ones that aren't treat AI as a business transformation, not a technology deployment.
Diagnosing which category you're in isn't complicated. Does your AI portfolio have financial objectives? Is it business-led or technology-coordinated? Are the pilots tied to value cases? The answers are usually quick, and usually uncomfortable.
The harder question is what to do once you've made that diagnosis.
Until next week,
Kendall -
Find me on LinkedIn or Book a 1:1 call
Not a subscriber yet? Subscribe here
