Ahead of Schedule | Your weekly AI edge for project leaders

Last week I shared 5 AI prompts for the day-to-day grind of PM work. Today I'm going deeper on the moment that separates good PMs from great ones: the project kickoff.

Most projects fail because of what doesn't happen in the first 30 days — the assumptions that go unquestioned, the scope that goes undefined, the stakeholders who get misaligned before they even know it.

Here's the exact process I use (with AI) to go from a messy project brief to a structured 30-60-90 day plan that everyone can actually execute.

The Problem with Traditional Kickoff Planning

The old approach: read the brief, open a blank doc, stare at it, create some version of a Gantt chart that's outdated by the end of Week 1.

The new approach: treat the brief as raw material, use AI to extract and structure the key information, then apply your judgment to fill the gaps.

Let me walk you through it.

Step 1: The Brief Deconstruction Prompt

Before you can plan, you need to understand what you've actually been handed. Most briefs are written by people who know what they want but not what it takes to get there.

The prompt:

"Here is a project brief: [paste brief]. Please extract and organize the following: (1) Primary objective in one sentence, (2) Key deliverables with a clear definition of done for each, (3) Stated constraints (budget, timeline, resources), (4) Implied constraints that aren't stated but are likely true given the context, (5) Stakeholders mentioned and their likely concerns, (6) What's notably absent from this brief that a PM would need to know."

What you're looking for: That last item — "what's absent" — is gold. AI has seen thousands of project briefs. It will surface the questions you need to ask before you even start.

Step 2: The Assumption Map

Every project runs on assumptions. The dangerous ones are the ones nobody has named.

The prompt:

"Based on this project scope: [paste scope], generate a list of the top 15 assumptions this project is making. Categorize each as: (A) Safe assumption — low risk if wrong, (B) Risky assumption — needs validation, (C) Critical assumption — project could fail if this is wrong. For each Category C assumption, suggest how to validate it in the first two weeks."

Run this with your team in a kickoff meeting. Watch how quickly it surfaces disagreements about things everyone thought were already decided.

Step 3: The 30-60-90 Generator

Now you have enough to build the plan. This prompt produces a structured first draft:

"I'm managing a [describe project]. The primary objective is [objective]. Key deliverables are: [list]. Key risks/assumptions we've identified: [list]. Team size: [N] people with these roles: [roles]. Please create a 30-60-90 day plan with the following structure:

Days 1-30 (Foundation): What needs to be true by end of month 1 for this project to succeed? List 5-7 specific outcomes.

Days 31-60 (Build): What are the key milestones in the execution phase?

Days 61-90 (Validate & Adjust): What should you be reviewing, testing, or adjusting?

For each phase, flag the top 2 risks and one early warning signal that would indicate the phase is going off track."

This produces a first draft in 60 seconds. Plan for 20-30 minutes of refinement with your team's input.

Step 4: The RACI Clarity Check

One more step that most PMs skip: making sure everyone knows who's doing what.

The prompt:

"Here are the key activities for my project: [list from your 30-60-90]. Here are the stakeholders: [list with roles]. For each activity, suggest a RACI assignment (Responsible, Accountable, Consulted, Informed). Flag any activities that have: no clear accountable owner, too many people marked Responsible (a common planning mistake), or gaps where no one is Consulted but probably should be."

The Real Value

The output of these prompts isn't the plan — it's the conversations the plan triggers.

When you walk into your kickoff meeting with a structured 30-60-90 day plan, a set of named assumptions, and clear RACI assignments, you change the meeting from "what should we do?" to "where do we disagree and why?"

That's a much better meeting to be in.

What's Next

Next week: The status update problem. I'll show you how to go from 45 minutes of writing a weekly update to 5 minutes — without sacrificing quality.

Subscribe link: buttondown.com/marcustillerman

Until next week,

Marcus Tillerman
Sr. Project Manager | NovaBridge Solutions
Columbus, Ohio

P.S. — If you tried any of last week's prompts, I'd love to hear how it went. Just hit reply.