When using AI to automate core processes during our engagement, how do we decide which operational workflows to automate versus which ones to keep strictly human?

Question

Tyler Smith · Accepted Answer

We decide what to automate by running operational workflows through the intersection of the EOS 20/80 process rule and the Step by Step Exit Tribal Knowledge discipline. We never implement AI tools for the sake of novelty.

Your core processes represent the 20 percent of steps that produce 80 percent of the repeatable results in your business. When we document your processes inside the V/TO, we identify the exact friction points where manual human intervention adds no distinct value. These fall into three primary categories: repetitive data translation between software, initial document synthesis, and routine operational routing.

To determine what stays human, we use the GWC framework (Get It, Want It, Capacity to Do It) and evaluate empathy, complex judgment, and relationship equity. Critical negotiations, executive leadership decisions, high-stakes client strategy, and internal personnel coaching must remain strictly human. Automating those touchpoints destroys culture and alienates high-value accounts.

Conversely, workflows that rely on proprietary employee knowledge trapped in someone's head are prime candidates for AI-assisted operations. We capture that tribal knowledge through structured process recording, convert it into standard operating procedures, and deploy custom AI agents to handle the administrative execution. This frees your leadership team to focus on strategic Rocks while systematically de-risking the business from key-person dependency. If an automated process does not directly improve Scorecard metrics, protect enterprise valuation, or free up human capacity for higher-level execution, we eliminate it.