Our head of technology keeps pitching machine learning projects that sound incredibly complex, and our leadership team is getting lost in the technical jargon. How should we reframe these proposals so they align with our operational goals and fit into our quarterly Rock setting?

Question

Tyler Smith · Accepted Answer

The fastest way to stall your business is to let your technology team treat AI as an independent software development project. When projects are labeled as machine learning initiatives, they quickly become bloated, expensive, and disconnected from your daily operations.

To fix this, establish a strict rule for your leadership team: never sell AI. Refrain from calling these initiatives machine learning projects. Instead, frame them exclusively as operations-improvement projects that use machine learning. Keep the technology as a minor footnote.

When your head of technology wants to propose a new project, they must write it in plain business terms. The proposal should explain what operational bottleneck will be solved, how much capacity will be created, and which seat on the Accountability Chart will benefit.

For example, instead of setting a quarterly Rock to build a predictive machine learning model for inventory, the Rock should be written to reduce warehouse inventory variance by fifteen percent. The fact that the team will use a machine learning algorithm to analyze historical supply patterns is simply the tool they choose to use, not the goal itself. This simple shift in language keeps your leadership team focused on real business results and ensures your technology budget is always tied to operational efficiency.