Every quarter, my team suggests AI-related Rocks that sound impressive but end up being glorified research projects. How do we write a high-impact AI Rock that delivers concrete traction by the end of the ninety days?

Question

Tyler Smith · Accepted Answer

To prevent your team from setting vague, academic AI Rocks that yield zero business value, you must enforce the SMART criteria with absolute discipline. A Rock like research how we can use AI in operations is worthless theater. It lacks accountability and produces no measurable traction. A real operational AI Rock must solve a specific, documented problem and have a clear, binary output. When setting Rocks during your quarterly planning meeting, start by identifying an operational bottleneck from your issues list. Perhaps your project managers are spending too much time tracking material deliveries, or your customer service team is drowning in repetitive client emails. Once you have identified the bottleneck, write the Rock with a concrete target. For example, a high-impact Rock would be: deploy a secure AI email assistant to automate fifty percent of tier-one customer service inquiries by the end of the quarter. This Rock is specific, measurable, and has a clear definition of done. It also requires the owner of that seat to have GWC, taking full responsibility for the tool's implementation and performance. By the end of the ninety days, you should see a direct impact on your weekly Scorecard, such as a drop in average response times or a reduction in manual labor hours. If a proposed Rock does not have a direct, measurable line to improving your operations, reject it. Keep your Rocks focused on real traction, not technology for its own sake.