The CPO 30/60/90

The 90 days are an audit, not a tour

Most first-90-days plans for a new CPO are politeness schedules. Meet everyone. Listen a lot. Find a quick win. Don't break anything. I have run that plan. It produces a leader who is well liked at day 90 and operating on borrowed conviction at day 180, steering with a model of the company assembled entirely from what people chose to tell them.

I have stepped into the product leadership seat several times now, at Commure, Crisis Text Line, SOCi, and Smartcat. The plan below is what survived. It treats the first 90 days as an audit with a deliverable, not a tour with a vibe. The deliverable is a day-90 readout your board will quote back to you for a year. Everything before it exists to make that document true.

This chapter is the operational version. For the argument about why the classic plan expired, see The First 90 Days as an AI-Native CPO(coming Jul 2). For the scoreboard you inherit the moment you take the job, see the CPO Mandate.

The short version

The CPO 30/60/90 is built around four audits and three artifacts. Days 1 to 30: run the reality audit, the money audit, the quality audit, and the decision audit, while logging every commitment you make in a trust ledger. Days 31 to 60: instrument what the audits exposed, build your coalition map, and ship one visible decision that signals the new bar. Days 61 to 90: write the kill list, place your first two or three bets, and deliver a day-90 readout structured as an SCQA memo to the exec team and board. Do not reorg, do not rewrite strategy in week 2, and do not spend the 90 days collecting opinions you cannot falsify. The templates for the ledger, the interview guide, and the readout are in the CPO First-90 Kit(coming Jun 10).

Day zero: before you walk in

The highest-leverage hours of your first 90 days happen before day 1. Three moves.

Become a customer. Sign up for the product with a personal email. Hit the paywall. File a support ticket. Try to cancel. You will never again experience the product without the org's framing wrapped around it. This is your only unanchored read. Take notes you can quote later.

Read the unguarded documents. The last four board decks, the last two postmortems, the public changelog, the pricing page history on the Wayback Machine. Decks tell you the story the company tells itself. Postmortems tell you the truth it admits under pressure. The gap between them is your first map of the org.

Write your priors down as falsifiable statements. Not "I think the team is strong" but "I believe the activation drop is a pricing problem, not an onboarding problem, and I expect the data to show X." Ten of these. Date the file. The first 30 days are designed to break these statements, and you cannot notice your model updating if you never recorded the model. I learned this one late. It changed how I onboard more than anything else on this page.

Days 1 to 30: the four audits

The first month is four audits run in parallel, plus one discipline that protects your credibility while you run them.

The reality audit. Use the product on real tasks weekly, not as a demo. Sit in at least eight raw customer calls, unedited, before you read anyone's synthesis. The synthesis is downstream of someone's judgment about what mattered, and you do not yet know whose judgment to trust. Build your own read first. My interview guide works as well for this as it does for discovery.

The money audit. Get cost per outcome visible by workflow. Pull compute and agent spend out of the aggregate cloud bill and attribute it to the workflows that incur it. In every AI-era product org I have looked at, at least one workflow was quietly underwater and nobody owned the number. Finding it pays for your first quarter. The full method is in Gross Margin Is Your Job Now.

The quality audit. List every production AI feature. For each, ask one question: what is the eval score, and which way is it trending? The answers sort into three piles. There is a number and a trend (rare, treasure these teams). There is a number, somewhere, probably (the actual state of most orgs). Silence. The size of the third pile is your real quality roadmap. See The Eval Is the Spec for what good looks like.

The decision audit. This is the one nobody runs. Pick the last ten significant product decisions: a killed feature, a pricing change, a big bet, a deprecation. For each, reconstruct how it actually got made. Who raised it, where it was debated, who decided, how long it took, what evidence was in the room. You are mapping the real operating model, which is never the one on the wiki. Most of what you will want to change at day 60 lives in this map, not in the roadmap.

The discipline: keep a trust ledger. From hour one, log every commitment you make. "I'll look into that." "Let's revisit in two weeks." "I'll get you an answer." What you said, to whom, by when, status. New executives bleed credibility through small dropped promises they never noticed making, and the people watching you in month one are counting. Review the ledger every Friday. Close everything or renegotiate it explicitly. Trust at day 90 is compound interest on this file. The template is in the CPO First-90 Kit(coming Jun 10).

One thing you are not doing in month one: sharing conclusions. You are allowed observations and questions. The moment you broadcast a thesis, every subsequent conversation becomes a referendum on it, and your information supply gets curated to match. Hold the thesis. Collect the evidence. The deep dive on this month is the CPO listening tour, rebuilt as a ledger(coming Jun 8).

Days 31 to 60: instrument and align

Month two converts the audits into instruments and the conversations into a map.

Stand up the instruments. Whatever the money and quality audits exposed, make it permanently visible. Cost per outcome by workflow on a page. Eval scores with trend lines on every production AI feature. A direction metric that says whether the product is getting better, not just bigger. This is the difference between knowing the number once and leading with it. Strategy From Signals covers what to instrument; Direction Metrics covers how to choose the headline number.

Build the coalition map. By day 45 you know who runs on evidence, who runs on narrative, who is exhausted, who is dangerous, and who has been waiting years for someone to fix the thing they care about. Write it down as an actual map: allies, persuadables, blockers, and what each one needs to move. Your first bets will live or die on this map, not on their merits. I wrote the long version in the CPO coalition map essay.

Ship one visible decision. Not a quick win. Quick wins are usually small enough to be ignored and cosmetic enough to be resented. Ship a decision that signals the new bar. Kill one zombie initiative everyone privately agrees is dead. Publish the eval scores nobody wanted on a wall. Replace the status meeting with a decision-forcing review(coming Jun 16). The point is not the object, it is the precedent: decisions here now get made on evidence, in the open, and they stick.

Renegotiate the scoreboard. Somewhere in month two, your CEO's patience for "still learning" runs out, usually unannounced. Get ahead of it. Bring a draft of what you want to be measured on for the next year, anchored in the audit findings. Margin-aware, eval-backed, outcome-shaped. This conversation is much easier to have at day 50 with evidence than at day 100 with excuses. The mechanics are in days 31 to 60, instrumented(coming Jun 15).

Days 61 to 90: the kill list, the bets, the readout

Month three is where you spend the credibility you banked.

Write the kill list. Every product org carries initiatives, ceremonies, and features that survive on inertia. The audits told you which ones. List them, with what each costs and what killing it frees up. You will not kill them all by day 90. You will kill two, publicly, with a written explanation. The rest go in the readout as commitments. The craft of killing well is in The Deprecation Playbook and the killed-feature playbook.

Place two or three bets, and name the anti-bets. Your first bets should be few, evidence-backed, and reversible-aware: know which ones you can walk back and which you cannot, and weight the irreversible ones accordingly. Just as important, name what you are explicitly not doing. An anti-bet list is the cheapest alignment tool that exists, because it converts a hundred future hallway debates into one paragraph. The Anti-Backlog is the standing version of this practice.

Deliver the day-90 readout. One memo, SCQA-shaped, five sections: where the product actually is, the two or three constraints that matter, the bets and anti-bets, the kill list, and the scoreboard you want to be judged on. Written to be read, not presented. Walk the exec team through it, then the board. This document is the actual deliverable of your first 90 days, and it is the artifact people will quote back to you for a year, so write it like you mean every sentence. Structure and template are in the strategy memo template(coming Jun 17) and the board-facing version in Investor and Board Narrative. The full month is broken down in days 61 to 90, first bets(coming Jun 24).

What not to do

Do not reorg before day 90. You would be restructuring a system you do not understand using a model assembled from interviews with people who each had an agenda. Two exceptions: an integrity or safety problem, and the leader everyone is waiting for you to deal with, where every week of delay reads as endorsement. Otherwise decide the operating model first and let structure follow. The Product Operating Model is the decision that comes before any org chart.

Do not rewrite the strategy in week 2. Whatever document you produce that early is your priors wearing the company's letterhead.

Do not run the listening tour as theater. Forty meetings that produce a word cloud is a month of evidence spoiled. Run it as a ledger with structured questions and falsifiable claims, or do not bother.

Do not skip the trust ledger because it feels bureaucratic. It takes four minutes a day. It is the cheapest insurance you will ever buy on your own reputation.

The throughline

Everything above serves one shift. The classic CPO arrived to optimize an execution machine, because execution was scarce. Execution is cheap now. What is scarce is an accurate model of reality, instruments that keep it accurate, and the judgment to bet on it. The first 90 days are where you either build that or substitute charm for it. Charm decays. Instruments compound.

Pick one thing this week

If you are about to start the role: write your ten priors as falsifiable statements and date the file. It costs an evening and it will reshape your entire first quarter.

If you are already in the seat, whatever the day count: start the trust ledger today, backfilled with every open commitment you can remember. Then ask the eval question of your most important AI feature and watch what the org does with it.

The templates for everything in this chapter, the interview guide, the trust ledger, the coalition map worksheet, and the day-90 readout skeleton, are in the CPO First-90 Kit(coming Jun 10).

Sources: Michael Watkins, The First 90 Days for the classic transition frame this chapter argues with, Barbara Minto, The Pyramid Principle for the SCQA structure of the readout, Marty Cagan / SVPG on product leadership transitions.