Your First 30 Days as a PM: Build a Signal Map, Skip the Coffee Chats

The standard first month of a new PM job looks like this: thirty coffee chats, the wiki, the all-hands recordings, a notebook full of org charts. At day 30 you know everyone's name and nothing about how the org actually decides. You built relationships. You did not build a model.

I have started over enough times to know which one matters. Relationships follow usefulness. A model is what makes you useful. So I spend the first 30 days on two things the default plan skips entirely: decision archaeology and a signal map. This post is the mechanics. The full 90-day arc lives in The PM 30/60/90.

The short version

Your first 30 days as a PM should produce a model of the org, not a contact list. Three moves: run decision archaeology on the last 90 days of your product area (five places to dig, five questions per decision, an agent does most of the work in an afternoon), build a signal map of the seven places truth enters the building and audit whether each gets captured, synthesized, and routed, and write a manager contract in week one that includes the question "what are you silently worried about." Hold all opinions until day 31. Templates are in the PM First-90 Kit.

Decision archaeology: the method

The pillar chapter explains why archaeology works. Here is how to actually run it.

Where to dig. Five sites, in this order:

1. Merged PRs. Filter your area's repo to the last 90 days, merged only. You are not reading code. You are reading PR titles, descriptions, and the arguments in review threads. The fights in PR comments are the org's real engineering values, written down.
2. Launch posts. Internal announcements, changelogs, the #launches channel. What got celebrated tells you what gets rewarded.
3. Killed projects. Harder to find because nobody announces a kill. Search Slack for "deprioritize," "pause," "revisit next quarter," "park this." Ask one engineer: "what got cancelled this year?" They always know.
4. Escalation threads. Search for "urgent," "escalation," the CEO's name in your product channels. Escalations show you the org under stress, which is the only honest version of any org.
5. Pricing and packaging changes. Even small ones. Pricing decisions cross more org boundaries than any other decision type, so one pricing change teaches you the whole decision topology.

The five questions. For each significant decision you find, answer:

1. Who raised it first?
2. Where was it debated? (A doc, a meeting, a DM thread, nowhere?)
3. Who actually decided? (Often not the person with the title.)
4. What evidence was in the room? (Data, a customer quote, a vibe, an exec preference?)
5. What happened after? (Shipped, stalled, quietly reversed?)

The afternoon version. This used to take three weeks of lurking. With an agent it takes an afternoon. Point a synthesis agent at the exports: PR list, the launches channel history, the escalation threads you found. Ask it to build a timeline of the last 90 days with the five questions answered per decision and confidence flags where it is guessing. Then spend two hours verifying the flagged items by asking humans. The setup is the same scaffolding described in The PM Agent Stack, and you should have built it before day one.

The output is one document: a 90-day decision timeline. By week three you will be the best-informed person in most meetings about recent history, which is a strange and useful kind of credibility for someone with no tenure.

The signal map worksheet

A stakeholder map tells you who has power. A signal map tells you where truth enters the building. In every company I have worked at, truth enters through roughly seven doors:

1. The support queue
2. Sales call recordings
3. Churn surveys and exit interviews
4. Product analytics
5. The community or customer Slack channel
6. The engineer who quietly knows everything
7. The exec who actually talks to customers

For each of the seven, answer four questions:

The pattern you will find, almost everywhere: signal is captured but not synthesized, or synthesized but routed to someone who cannot act on it. Sales calls are usually the biggest pile of unprocessed truth in the building. That gap is your month-two artifact waiting to happen, and the 20-minute customer call triage pipeline is how you will process it.

Two of the seven are people, not systems. Find the engineer who knows everything by asking three engineers "who would you ask about why things are the way they are?" The name that comes back twice is your person. Find the customer-facing exec by checking who actually shows up in call recordings. Buy both of them lunch. These are the only two "coffee chats" I consider mandatory.

Then subscribe to everything. Route what you can into a daily digest, the way Continuous Listening lays out. Long-term, your job is to sit closer to the signal than anyone else in the org. Month one is when you claim the seat.

The manager contract conversation

Week one, book an hour with your manager and say you want to write down the operating agreement. Ask these, verbatim:

1. "What do I own, and what do I explicitly not own?"
2. "What does good look like at day 90? Describe the version of me you would brag about."
3. "How do decisions get made between us? What do you want to approve, see, or only hear about after?"
4. "What has burned people in this role before?"
5. "What are you silently worried about?"

The last question is the one almost nobody asks, and it is the whole conversation. Managers carry a private worry list about every new hire and every product area. The product they suspect is in trouble. The stakeholder they cannot manage. The commitment they regret. Asking directly gets you the real job description, and it tells your manager you can handle uncomfortable information, which is its own signal.

Write the answers as one page. Send it back: "here is what I heard, edit anything I got wrong." Now expectations are an artifact instead of a vibe, and at day 90 you will close the loop against it.

The daily capture habit

All of this produces a firehose, and the first month of a new job is the highest-value input stream of your entire tenure. Most PMs let it run straight to the floor.

Ten minutes at the end of every day: what did I learn, what surprised me, what question opened up. Into the second brain you set up before day one (the PM second brain is my setup). The surprises matter most. In six weeks nothing will surprise you anymore, because you will have a story too. The capture file is the only record of what the org looked like before you joined its consensus.

A worked week one

Generic and sanitized, but this is the shape:

Monday: laptop, accounts, use the product as a customer one more time with internal eyes. Tuesday: pull the PR list and launches channel, hand them to the agent, get the draft timeline by evening. Wednesday: manager contract conversation, write the one-pager. Thursday: chase the three killed projects the agent flagged, lunch with the engineer-who-knows. Friday: first pass at the signal map table, subscribe to the support queue and the call recording library, end-of-day capture, done by five.

That is a real week. No heroics. By Friday you have more model than the coffee-chat plan produces in a month.

Failure modes

Performing curiosity instead of practicing it. Asking questions in meetings to be seen asking questions. People can tell. The test: does the answer change what you do next? If you are not writing answers down and following the threads, you are performing.

The premature opinion. Around day 25 the pressure to have a take becomes physical. Someone asks "so, what do you think we should do?" in front of the room. Hold. The honest answer is a script: "I have a hypothesis but I want two more weeks of evidence before I put my name on it. Here is what I have noticed so far," and then share an observation, not a recommendation. An opinion at day 25 is a guess wearing your reputation.

Archaeology becoming gossip. The five questions surface bodies. Decisions that were reversed quietly, evidence that was ignored, a kill that embarrassed someone. The moment you narrate any of it as "who screwed up," the archaeology turns into politics and your sources dry up. The frame, always: "I am trying to learn how we decide, not relitigate what we decided." Say that sentence out loud when you ask about killed projects. It matters.

Pick one thing this week

If you are inside your first 30 days: pull the last quarter's merged PRs and launch posts tonight and start the timeline. Two hours with an agent, and the rest of the month gets easier.

If you start soon: set up the second brain and the synthesis agent before day one. No company access needed.

If you are years into a job and never built the signal map: build it anyway. The seven-source audit works at any tenure, and the gaps it exposes are next month's artifact. Then read PM Days 31 to 60: Ship One Artifact That Compounds(coming Jun 22) for what to do with them.

Sources: Michael Watkins, The First 90 Days for the classic onboarding frame this replaces parts of, Teresa Torres, Continuous Discovery Habits for the interview and signal cadence, Julie Zhuo, The Making of a Manager on making expectations explicit.

Further reading

• The PM 30/60/90, the full 90-day arc this post expands
• Continuous Listening, where the signal map feeds
• The PM First-90 Kit, the contract and signal map templates