Skip to main content

Agent Opportunity Analysis

A function-by-function method that sorts a company's work into what to automate, what people should keep, and what should become an AI system, ranked by return.

What It Is​

Agent opportunity analysis is the disciplined way to decide where AI belongs in a business. Instead of asking the vague question "how do we use AI," it walks the company one function at a time, sales, operations, finance, delivery, recruiting, and for each one inventories the actual work being done. Every piece of that work then gets sorted into one of three buckets:

  • Automate it. Repetitive, rule-bound, high-volume tasks where a system does the whole thing and a person checks the exceptions.
  • Keep people on it. Judgment, relationship, taste, and accountability work where a human should stay in the seat and AI, at most, assists.
  • Turn it into an AI system. Work that should be rebuilt around AI as a standing capability the company owns, not a one-off task and not a job left untouched.

The third bucket is where most of the value lives and where most amateurs never look. It is not automating a task or leaving a job alone. It is redesigning how a function works so that an AI system carries the load with people supervising it.

Why It Is the Analytical Core​

This sorting exercise is the engine of any honest AI roadmap, which is also an exit-readiness roadmap. Without it, AI spend is scattershot: a chatbot here, a copilot there, no theory of where the return comes from. With it, every proposed change is tied to a function, a bucket, and a number.

That number is the discipline. Each opportunity is ranked by ROI, so the company does the highest-return work first and can say no to the low-return work that vendors love to sell. The ranking is what turns a wish list into a plan, and it is what an owner can carry to a board: here are the functions, here is what we are changing in each, here is the expected return, here is the order. It is the structured case the transformation flood leaves owners unable to build on their own.

What It Depends On and Produces​

Agent opportunity analysis cannot run on a company that has not written itself down. To sort the work in a function, you first have to see the work, which means the processes have to be documented. This is why documentation equals transferability is the precondition: the same documented truth that makes a business transferable is what makes its work legible enough to analyze for AI. The analysis reads from, and writes back into, the organizational truth repo.

Done well, the output is a ranked, function-by-function roadmap: a concrete list of what to automate, what to leave with people, and what to rebuild as an AI system, ordered by return. Executed, that roadmap is what actually earns the AI valuation premium, because it is the difference between a company that bought some AI tools and one whose operations have genuinely changed. It also doubles as exit-readiness work, since most of what it produces, documented process and reduced owner dependence, is exactly what a buyer pays up for.

Further Reading​

Sources: Built for Exit, Supersuit Up or Get Left Behind; Snider, Walking to Destiny (function-level value work).