In M0-S2, you learned the 5 mental models that govern AI-first PMs. Two of them are about to collide in this session.

Model 3 — "The PM's Judgment Is the Moat" — says that AI can compress admin but cannot make trade-off decisions. The same logic applies to AI tools themselves. You can't evaluate tools by watching demos or scanning ProductHunt reviews. You need a framework that forces YOU — the PM — to apply judgment systematically. That's the scorecard.

Model 4 — "Quality Gates Are for AI Too" — says AI output deserves the same review rigor as human work. Tools are no different. Before you adopt any tool into your workflow, it needs to pass a gate. The 5-factor scorecard IS that gate.

And from your M0-S2 task map: recall which of your tasks you classified as AI-feasible. That classification is your north star. A tool is only worth evaluating if it maps to a task you actually want automated. If a tool doesn't address your feedback synthesis pain point, its demo quality doesn't matter.

You built a system in M0-S1 (your weekly review Project). You classified your work in M0-S2. Now you're building the filter that protects your time from the firehose of new AI tools. This session is Model 4 applied to the market itself.

---

Every week there's a new AI tool for PMs. One promises auto-generated PRDs. Another promises "customer insight at scale." A third claims to automate your entire product workflow. Your LinkedIn feed is a firehose of demos. Every founder says their tool is "the one."

You cannot evaluate them all. You need a framework that tells you, in 15 minutes, whether a tool deserves more of your time or should be ignored until it matures. This session gives you that framework — the same one you'll use for every AI tool you encounter for the rest of your career.

---

For every AI tool you evaluate, answer these 5 questions. Not "is it cool?" Not "do other PMs use it?" These five:

1. Problem Fit — Does it solve YOUR actual problem?

What to ask: What does this tool claim to solve? Is that a real pain point in MY work — or a solution looking for a problem?

Red flags:

• The tool solves a problem you don't have ("AI-powered roadmap visualization" — you use a Google Sheet and it works fine)
• The tool claims to replace PM judgment ("AI decides what to build next")
• The tool description is vague about what it actually does ("revolutionize your workflow")

Green flags:

• You read the description and think "this is exactly the thing that takes me 3 hours every Friday"
• The tool has a specific, narrow use case (not "AI for PMs" but "AI that generates stakeholder updates from Jira data")

2. Workflow — How many steps from open to useful output?

What to ask: If I signed up right now, how many minutes until I have a useful output I'd actually use? Not a demo. Not a tutorial. MY work output.

Red flags:

• Requires setup beyond account creation (API keys, configuration files, integrations)
• No free tier — you can't test without paying
• "Book a demo" instead of "Try it now"
• Requires CLI, terminal, Python, or coding
• The onboarding tutorial is longer than 10 minutes

Green flags:

• Sign up with Google → upload a file → get output in under 10 minutes
• Free tier that lets you test with real (not sample) data
• Web UI — no installation, no configuration, no terminal

3. Setup Cost — What's the time investment to get to first useful output?

What to ask: Accounting for account creation, learning the interface, and generating my first useful output — how much time? Is the time investment proportional to the time it'll save me?

Red flags:

• Setup > 30 minutes before you see any output
• Requires importing data from 3 different tools before it works
• Needs ongoing configuration (weekly maintenance, context re-upload)
• Enterprise sales process — "we'll get back to you with pricing"

Green flags:

• First useful output in under 10 minutes
• Works with files you already have (export from Jira, upload a PDF, paste text)
• Remembers your context (you don't re-upload every time)

4. Maintenance Burden — What's the ongoing cost?

What to ask: After I set this up, what do I have to do every week/month to keep it working?

Red flags:

• Re-upload context every session
• Manually trigger every run (no scheduling)
• Output quality degrades over time unless you tweak settings
• Expensive — > $50/month for individual use without clear ROI

Green flags:

• "Set it once, it runs on schedule" (n8n, Claude Projects)
• Context persists across sessions
• Pricing is transparent and proportional to value
• Free tier covers the core workflow; paid is for power users

5. Integration — Does it work with my actual toolchain?

What to ask: Does this tool talk to Jira, Slack, Notion, Google Workspace, and/or Salesforce? Or is it a walled garden that requires me to work inside it?

Red flags:

• "Import your data" — but only supports CSV upload, not live integrations
• Creates a separate silo you have to maintain alongside your existing tools
• No export — output lives inside the tool, can't be sent to Slack/email/Notion

Green flags:

• Reads from your existing tools (Jira API, Google Drive, Slack)
• Output can be emailed, posted to Slack, or saved to Notion
• Works alongside your toolchain, doesn't try to replace it

---

After answering all 5 factors, assign one of three verdicts.

✅ Adopt

• Scores well on all 5 factors
• Solves a real pain point you have TODAY
• Takes < 15 minutes to first useful output
• Free or affordable
• Integrates with your toolchain

Action: Start using it this week. Incorporate into your workflow. Come back in 2 weeks and re-evaluate: is it saving the time you expected?

👀 Watch

• Promising but has 1-2 red flags
• Might be right for you in 3-6 months (missing integration, too expensive, too new)
• Solves a real problem but the execution isn't there yet

Action: Bookmark it. Set a calendar reminder for 3 months. Check if the red flags have been resolved. Don't spend more time on it now.

❌ Pass

• Multiple red flags
• Requires CLI, coding, or technical setup beyond your comfort level
• Claims to do things AI can't actually do (strategy, decision-making, judgment replacement)
• Solves a problem you don't have

Action: Move on. Don't feel FOMO. There will be 10 more tools next month. Your time is better spent mastering the tools you've already adopted than chasing every new launch.

---

Tool 1: Claude (You Already Use It)

Apply the 5-factor scorecard to Claude based on your M0-S1 experience. Be honest — critique where warranted.

Tool 2: Pick a Tool You've Been Curious About

Choose one AI tool you've seen mentioned, bookmarked, or been "meaning to try." Options if you don't have one:

• NotebookLM (notebooklm.google.com) — Google's free RAG tool, upcoming in Module 2
• Perplexity (perplexity.ai) — AI search with citations, good for competitive research
• Granola (granola.ai) — AI meeting notes, if you haven't tried one yet

---

You now have a decision framework for every AI tool you'll encounter. The landscape changes monthly. The framework doesn't.

When a colleague sends you a link with "you have to try this AI tool," you have two choices:

1. Drop everything, sign up, spend 30 minutes exploring, realize it doesn't solve your problem
2. Apply the 5-factor scorecard in 5 minutes, assign a verdict, move on with your day

Option 2 is how AI-first PMs operate. They don't chase every tool. They evaluate systematically, adopt what works, watch what's promising, and pass on the rest.

---

Save your scorecard template. Use it every time a new AI tool crosses your radar. In 6 months, you'll have a personal library of evaluated tools — adopted, watched, and passed — and you'll have wasted zero hours on demos that go nowhere.