Lab: The CEO Brief
Stryg for at vise menuen
No more scaffolding. No worked example. You've learned the moves – this chapter is where you prove they're yours.
The Scenario
It's Sunday evening. Your CEO sends you a message:
Quick favor – for tomorrow's leadership offsite, can you put together a read on where we stand vs the top 3 competitors? Just the highlights – what we win on, what we lose on, biggest risk.
You don't have an hour or two. You have a Competitive Research Analyst agent – the one you're about to build.
What the Leadership Team Needs
They want a brief that takes a position:
| What They Need | What That Means |
|---|---|
| A win they didn't already know about | Named with evidence |
| A loss they're underestimating | Named with evidence |
| A risk they should be losing sleep over | Named in plain English |
Self-Evaluation Rubric
| Test | Ask Yourself |
|---|---|
| Specific or generic? | Does every claim point at a number, a competitor name, a date? |
| Position or summary? | Is there a thing the agent thinks, or just a list of things that exist? |
| Sources or invented? | Can you trace every fact back to one of the tools? |
| Forwardable? | Would you paste this into Slack and tag your CEO, or edit it first? |
Task
Analyse our competitive position against the top 3 PM SaaS competitors. Where do we win, where do we lose, and what's our biggest strategic risk for next quarter?
What You've Actually Built
Not a chatbot. A teammate. The same agent shape – persona, tools, structured output – works for triage, metrics readouts, prospect research, post-mortems, board updates.
Takeaway: you're not learning to use AI. You're learning to assemble it.
Where This Goes Next
Same persona. Same tool routing. Same output format. Real data underneath – the moment those tools connect to your real Notion, metrics, and competitors.
Tak for dine kommentarer!
Spørg AI
Spørg AI
Spørg om hvad som helst eller prøv et af de foreslåede spørgsmål for at starte vores chat
