Are investments involving vendor integration being funded through OpEx or CapEx? How are others approaching this?

I agree with the concerns about cost. The industry’s push for cloud-first and SaaS models was supposed to reduce infrastructure costs, but those costs have not decreased as expected and may even be increasing. As a result, some organizations are reconsidering on-premises solutions for certain applications. We’ve learned from the cloud-first strategy not to put all AI investment in one area just because a vendor has an AI strategy. If upgrading to an AI-powered system reduces support staff from 25 to 15, that’s worthwhile; if it increases staff, there’s no gain. We need tangible returns before making major investments.

When it comes to existing vendors incorporating AI, I am concerned about the recurring increased spend due to the expense of implementation. For example, one vendor we use for surveying is integrating AI capabilities, but we’re not sure it’s valuable for us yet, and there’s a significant cost difference. They allow us to use their existing platform for another year before moving to the new AI-enabled model, which is much more expensive. Most of our AI-related expenses, such as Copilot, are ongoing subscription costs (OpEx) with a long tail, rather than CapEx. Much of our infrastructure is in AWS or Azure, and we’re still figuring out how to plan for these inevitable changes.