The 60-second update

What you say
in 60 seconds.

Say what shipped and what landed in one paragraph, with a source behind each number, and you earn the next four minutes of board time.

The frontier test

Is your strategy keeping up with what the models can now do?

Your product strategy got set when one model was the state of the art. Two better ones have shipped since. What the models can do moved twice this quarter alone.

If the strategy has not moved with it, you are planning against last quarter's tools. If it moves every few weeks, you are rewriting the plan without ever committing to one. The useful check is simple: name the last time the strategy changed because the tools changed, name who made that call, and name the week it happened.

The design test

Can your design team say what AI-native looks like in your product, in plain words?

“The screen builds itself around the question you asked.” Specific. “The agent does the work in front of you and you approve the step.” Specific. “AI-powered insights.” A label on the old screen. The difference shows up in what gets shipped, in how the board reads the screenshots, and in whether the team can explain the choice in under 30 seconds.

The read is whether design changed how the product behaves, or only the wording around it.

The quality-debt test

How much of your code do agents write, and does the checking keep pace?

Agents wrote 60% of your code this quarter, up from 12% six months ago. If the tests and checks grew at the same rate, you are covered. If they did not, you are taking on debt that never appears as a line item.

The board will ask the first half. Walk in with the second half already answered.

The org-shape test

Does the shape of the team still match the work?

The mix of product, engineering, and design was set for the work you did three years ago. The work has changed twice since.

When agents triple how much engineering gets through, and design goes from sketch to working screen in a day, the mix that fit then spends the budget badly now. The point is not a new ratio. It is that hiring follows evidence about where the work actually is, not the last quarterly doc.

◆The structure
[What shipped, with a number] · [Whether it landed, with a number] · [The gap between the two, and who owns closing it next quarter]

Three sentences. Both numbers. One gap, one owner. Under 60 seconds at a normal speaking pace. Nothing about transformation programs, swim lanes, or initiative health.

The readIllustrative, not a real quarter
Agent-led deploys
Moved the metric it was aimed at.
landed
Onboarding revamp
Climbing, not yet at the bar.
watch
Self-serve upgrade
Shipped. Nothing moved.
stalled

Three rows is the whole slide. The one in red is what the next four minutes are about, and somebody in the room owns it by name.

◆What to say at each stage

The story changes at each stage. The structure doesn’t.

React · the team is catching up

We shipped 30% more this quarter than last. Less of it landed than we planned, so what the team can do is running ahead of what customers feel. Closing that gap is next quarter. Here is the first move, and here is who owns it.

Augment · agents are in the loop

Cycle time came down 22% across all six functions, not only engineering. A customer signal now reaches a roadmap change in 11 days, down from 6 weeks. Both numbers moved because we changed how decisions get made, not because we ran a transformation program.

Orchestrate · the work is coordinated

Cost per shipped outcome is down 18%, with agents doing the first pass on research and drafts. Three of four bets paid off. We stopped the fourth in week 4 instead of month 9. Fewer calls, better ones, and every one of them has a name against it.

Lead · the team sets the pace

Each feature costs less to run than the one before it, and margin holds at ten times current usage. AI-native features are 64% of revenue. We spend 1.4 times more effort choosing the work than building it. The loop is running. Each quarter starts from what the last one taught us.

Compound · every cycle makes the next one cheaper

Two of our three competitors took over a year to match our first-quarter release. Net revenue retention is 134%. Customers who arrive through an agent are 28% of new business. The advantage is the loop itself, not any one feature on the roadmap.
◆The evidence chain

Every number has a source. Nobody has to take your word for it.

Each line in the examples above comes from somewhere. That is the whole discipline: before you say a number out loud to a board, know what produced it and who owns it. The pairing below is the version I use.

Cycle time came down
How the operations score moved over the last four cycles
The team is ahead of the product
How the team runs, read next to what it ships, cycle over cycle
Time from customer signal to roadmap change
Feedback loop quality, traced across every tool the signal passed through
Cost per shipped outcome
Cost per outcome and model spend, measured per feature
How many bets paid off
Decision quality, with the date each call was made and the person who made it
Share of revenue from AI-native features
Product assessment score, read next to revenue tagged by feature
Effort spent choosing the work vs building it
Research and discovery, read next to delivery velocity
Customers who arrived through an agent
Go-to-market function · adoption and expansion · signals tagged by source

If you also run the team: Which rituals still need a human. If you are sizing up your own job: Hire for the PM job that exists now.