Can you share a specific example from your organization where an AI investment directly competed with another critical IT priority? What was the outcome, and what would you do differently?

I don’t have an example of AI investment directly competing with another IT priority. Our philosophy is that innovation and “keeping the lights on” are not mutually exclusive; we budget for emerging tech every year, and this year it happened to be AI. Our budgeting and investment allocation is based on which business group is most interested in investing first, and costs are allocated accordingly. This approach allows us to invest in new technology without putting other priorities on hold. While we’re using Copilot and seeing increased usage, we still consider it a pilot. Measuring return on investment is challenging at this stage; the benefits are mostly productivity gains, and there isn’t a direct dollar-for-dollar return yet.

Our AI journey is still in the test and learn phase. Currently, we’re viewing AI from an investment standpoint and receiving additional funding to support these initiatives, especially in the generative AI space. The biggest trade-off is with our investment in Copilot and other AI capabilities due to the cost. To prioritize AI, I’m offsetting some lifecycle management spend on physical hardware and delaying those investments. We’re evaluating if Copilot licenses make sense for certain users based on price and value, and the near-term impact is likely on lifecycle management, either on the end-user side or core infrastructure.

We focus on three areas: farmer, dealer, and employee. Every initiative is scored against these criteria, and those with high scores are prioritized. This approach helps us eliminate competing initiatives and align our priorities.

The challenge is that while we see value in AI, articulating it is difficult. At Ferro, we have significant structural gaps, so it’s hard to divert resources from addressing those gaps to pursue AI initiatives whose benefits are not easily quantifiable. Our approach has been to acquire a few Microsoft Copilot licenses, distribute them across business functions, and encourage experimentation. We’ve created communities to share experiences and identify where AI can be justified. We are also prototyping agents through Copilot Studio, especially in support areas, to demonstrate value before making larger investments.

We are early in our AI journey and lean towards a "buy" philosophy for AI, preferring readily available solutions. However, due to our low maturity in AI, other priorities often take precedence. For example, we wanted to bring automation to our warehouse using AI technology, but this initiative consistently took a back seat to larger projects like ERP transformation and AR process automation. These projects are easier to justify because we have established ROI templates, whereas AI projects are harder to quantify in terms of return. In challenging economic times, especially with industry-specific tariffs, it’s difficult to prioritize AI investments without clear ROI.