Summary

This prompt is built for fractional and consulting operators selling into mid-market companies that have already restructured around AI. It takes what you know about a real buyer and returns three artifacts: a map of the operating-model gap between the AI in place and the return the buyer was supposed to get, a one-line name for the single highest-leverage missing piece, and an opening line you can paste into your next conversation with that buyer.

The prompt is anchored on a 2026 NBER working paper (Yotzov, Barrero, Bloom et al, WP 34836) covering roughly 6,000 executives across the US, UK, Germany, and Australia. More than 80 percent of firms reported no measurable impact from AI on either employment or productivity over the past three years. The work that produces the return is the operating model around the AI, and almost nobody is doing it. This is the prompt that helps you see that gap inside one buyer and start a paid conversation about closing it.

When to Use

Run this before any discovery call into a company that restructured in the last two quarters. Run it before writing a one-page scope. Run it before sending the cold outreach. Run it whenever a buyer has spent money on AI but the conversation about what is actually working has not happened yet.

Do not run this when the buyer has not deployed AI. The prompt assumes there is already an operating-model gap to find. If there is no AI in place, you are selling a different conversation.

Best Input

The more specific you are, the more useful the output. Provide:

• The company name, size (revenue or headcount), and industry
• What AI tools or agents they have already deployed (be specific: which vendor, what scope, when)
• What the buyer wanted the AI to do (the original business case)
• What is actually happening right now (the gap between expectation and reality, even if your read is partial)
• The named person you would be talking to (title and what they are accountable for)
• Any signal you have on the workflow, training, handoff, or metrics layer the AI is supposed to sit inside

If you only have two or three of these, run it anyway. The prompt will flag what it does not have and tell you the next question to ask.

The Prompt

You are an operating-model diagnostic working with a fractional operator preparing for a conversation with a real buyer. Your job is to surface the gap between the AI the buyer has deployed and the return the buyer was paid to produce.

Here is what the operator knows about the buyer:

[paste your inputs from the Best Input section above]

Do the following in this order: