Turning organizational intelligence into action begins with one discipline: findings are not actions. Every tested conclusion from the listening tour, the Influence Graph, and the alignment work is placed into an Action Portfolio with four positions. Protect: what already works, named publicly and left alone. Act: a small set of confirmed, reversible, visible first moves. Build: the one or two foundational commitments that will take longer than a quarter. Watch: what stays under open observation, each item with a review date. A move earns the Act position only by passing the Three Gates (confirmed, reversible, ready), and the first ninety days unfold in three deliberate moves: show the portfolio, deliver the first proof, set the rhythm.
Every article in this series has ended with the same quiet promise: the understanding you are building will eventually have to earn its keep. The listening tour gathered what the organization knows. The mapping work showed how it really operates. The testing phase separated what recurred and survived checking from what merely felt true. If you have done that work, you now hold a picture of the business you can defend in front of the people who lived it.
This article is about the step that picture was for: deciding what to do. It is the moment the entire system has been building toward, and also the moment of greatest risk, because months of careful listening can be spent in a single week of careless action.
The answer this series gives is not a longer list of initiatives or a faster one. It is a different shape for the first plan altogether: a portfolio of deliberate commitments in which what you leave alone is decided as carefully as what you change, and in which waiting, done in the open, counts as a decision rather than a failure to make one.
The organization that told you the truth is watching to see what you do with it. The first plan is not a test of ambition. It is a test of whether the listening was real.
- Findings are not actions. The first discipline of the operating phase is separating what you now know from what you will do. Every tested finding lands somewhere deliberate; not every finding deserves an initiative.
- The first plan is a portfolio, not a task list. The Action Portfolio holds four positions: Protect what works, Act where the evidence is strongest, Build the one or two things that matter most, and Watch the rest openly, each with a review date.
- Three gates guard the first moves. A change earns immediate action only when the finding behind it is confirmed, the move is reversible if wrong, and the organization has the attention and capacity for it now.
- What you protect is announced, not assumed. Naming what will not change releases the bracing that builds up around every new leader, and it tells the people who built what works that it was seen.
- Ninety days, three moves. Show the portfolio, deliver the first proof, set the rhythm. By day ninety the transition should be over, because ordinary leadership has begun.
The Moment the Organization Is Watching For
By the time a new leader reaches this phase, the organization has been watching them listen for the better part of a quarter. People have shared candid views and answered the same questions their peers were asked. They know something is coming. The question each of them quietly carries is not whether things will change. It is what will change, whether it will be fair, and whether any of what they said mattered.
That makes the first plan a communication before it is a strategy. Whatever you announce will be read as the verdict on three months of candor. Act on things nobody recognizes, and the listening is retroactively revealed as theater. Act on everything at once, and the candor turns to bracing. Act on nothing, and understanding becomes the destination instead of the starting point.
The stakes are measurable. Michael Watkins’ research on leadership transitions, drawn from surveys of more than two hundred company presidents and CEOs, put the break-even point of a typical newly hired or promoted leader, the moment they have contributed as much value as they have consumed, at around 6.2 months. The first plan is the lever that bends that curve. A plan that holds shortens it. A plan that has to be walked back extends it, sometimes past the point where the appointment recovers.
The record on ambitious change explains why the careful version wins. In McKinsey’s global survey on transformation efforts, only 26 percent of executives said the transformations they knew best had succeeded at both improving performance and sustaining the improvement. Most organizational change fails not because the ideas were wrong but because the change asked more of the organization than it could give, faster than trust could carry it.
So the operating phase opens with a narrow question, and everything in this article serves it: of all the things you now understand, what deserves to change first, what deserves to be defended, and what deserves more time?
Findings Are Not Actions
Start with the discipline the whole phase rests on. Take everything the previous three phases produced and write it as findings: the patterns that recurred, the strengths everyone underuses, the friction everyone has learned to live with, the places where the leadership team discovered it was not aligned. Findings describe the organization. They carry no verbs.
Then, on a separate page, write the candidate actions: the things you are tempted to do. The separation sounds trivial. It is not. Held apart this way, two honest tests become possible that a mixed list quietly prevents.
First, every action must trace back to a finding. If a candidate move has no tested finding behind it, what is driving it is instinct, or precedent from a previous company, or the pressure to look decisive. Those are not automatically wrong, but they have not earned the confidence the last three months built, and they should not borrow it. An action that cannot name its finding waits until it can.
Second, no finding is allowed to vanish. Some will produce action now, some later, and some only vigilance, but each one lands somewhere you can point to. The people who surfaced those findings will be watching for them, and a finding that silently disappears tells its source that speaking up was wasted effort.
Findings describe the organization; actions change it. Keep them on separate pages long enough to ask two questions of every candidate move: which tested finding is behind this, and where does every other finding go if not here? The first question keeps instinct from wearing the costume of evidence. The second keeps the organization’s candor from disappearing into a drawer.
The Action Portfolio
With findings and actions separated, the question becomes how to hold the first plan. The instinct is a list: prioritized, numbered, long. Lists are the wrong shape for this moment. A list treats every item as the same kind of thing and implies that, given time, everything on it will be done. Neither is true in a transition, and pretending otherwise is how new leaders end up sponsoring eleven initiatives and finishing none.
The first plan should be held the way an investor holds capital: as a portfolio of deliberate positions, balanced against each other, with scarcity built in.
The Action Portfolio is the operating instrument of the Organizational Intelligence System’s final phase: the complete set of a new leader’s first commitments, held in four positions. Protect names what already works and will not change. Act holds a small number of confirmed, reversible, visible first moves. Build holds the one or two foundational commitments that will take longer than a quarter. Watch holds the findings that deserve continued observation, each with an owner and a review date. Every tested finding lands in exactly one position, so nothing is rushed and nothing is lost.
Protect: Decide What Will Not Change, and Say So
The portfolio is filled in an unusual order: the first position is the one with no change in it. Before announcing a single initiative, name what the listening showed to be working, and commit to leaving it alone.
This is not a courtesy. It solves a real problem. Around every new executive, an organization braces: work slows, decisions wait for the new direction, and the most capable people quietly consider their options. The longer the listening went on, the higher that tension has climbed. Naming what is protected releases it. The teams running what works stop defending it and go back to running it.
Protection also pays a debt. Somebody built the things that work, usually without much recognition, and the listening tour told you who. Saying “this stays, because it works, and I understand why” shows the organization a leader who can see quality, not just problems. Of everything in the portfolio, this position costs the least and buys the most.
Act: A Few First Moves, Chosen for Evidence
The Act position holds the visible changes of the first ninety days, and its power comes from what it excludes. Three moves, maybe four. Each one traces to a finding that recurred, survived the record, and was recognized by the people closest to the work. Each one can be undone at reasonable cost if it proves wrong. And each one will be visible enough that the organization notices it happened.
Choose them where the evidence is strongest, not where the annoyance is loudest. The best first moves share a signature: they were named independently across the organization, the fix is well understood by the people who live with the problem, and what was missing was a leader willing to decide. Moves like that demonstrate, faster than any speech, that what people said in those conversations went somewhere.
This is also where momentum is manufactured, deliberately. John Kotter’s study of more than one hundred companies attempting major change found that the successful ones did not hope for early results; they planned visible short-term wins and engineered them into the effort. The Act position is that principle applied to a transition: a small number of early proofs, chosen on evidence, sequenced to land inside the first ninety days.
Build: The One or Two Things That Matter Most
Some of what the listening surfaced cannot be fixed with a reversible ninety-day move: an operating model that no longer fits the strategy, a gap the whole leadership team discovered it shared, a capability the next three years depend on. These belong in the Build position, and the discipline is focus. One foundational commitment, two at the very most, chosen from the deepest patterns the evidence produced.
Build work starts inside the first ninety days and does not pretend it will finish there. It moves at the pace of deliberation: sequenced carefully, communicated honestly as long-term work, designed with the people it affects. Its early milestones are foundations rather than results, and saying so out loud protects the effort from the impatience that kills foundational work everywhere.
Watch: Keep Looking, On Purpose, In the Open
The last position holds everything that deserves neither action nor dismissal: the finding that surfaced only once but from a source who has been right before, the pattern that is real but not yet understood, the change that is right but arrives when the organization has no capacity left for it. These go under observation, engineered so the watching cannot decay into avoidance.
Every Watch item carries three things: an owner, a reason it is waiting, and a date it will be looked at again. Then it is said out loud, in the same readout as everything else: here is what I am not acting on yet, and here is why. That openness separates waiting as a decision from waiting as a hope that the problem resolves itself. The Watch position is not the plan’s leftovers. It is the patience that built the understanding, still working.
| Position | What belongs there | What the organization sees |
|---|---|---|
| Protect | What the evidence shows is working, and the people and practices behind it. | A public commitment that it stays, and credit to those who built it. |
| Act | Three or four moves that are confirmed, reversible, and visible. | Early changes that trace to what people actually said. |
| Build | One or two foundational commitments drawn from the deepest patterns. | Long-term work, named honestly as long-term work. |
| Watch | Real but unconfirmed findings, and right-but-not-yet changes. | An open list with owners and review dates, revisited on schedule. |
Three Gates Between a Finding and a First Move
The portfolio tells you where everything lands. The gates decide the hardest allocation: what earns the Act position. Three questions, asked of every candidate move, in order. They take a minute to ask, and they are worth asking of everything.
Is it confirmed? The first gate simply enforces the previous phase. The finding behind the move recurred across unrelated conversations, lives in the notes, was recognized by the people closest to the work, and squares with what customers and the numbers show. A move built on anything less is a guess, and guesses do not belong in the first ninety days, when every action is being read as a signal of how decisions will be made from now on.
Is it reversible? The second gate borrows the most useful decision distinction in modern management. Jeff Bezos, in his 2015 letter to Amazon shareholders, separated decisions into two types: one-way doors, consequential and nearly impossible to reverse, which deserve slow, careful deliberation, and two-way doors, which can be walked back through if they prove wrong, and which should be made quickly by people close to the work. Transitions invert the failure Bezos described. New leaders, eager to signal decisiveness, treat one-way doors like two-way doors: the reorganization announced in week two, the senior departure forced before the map is understood. The gate exists to catch exactly those. In the first ninety days, reversible moves may go fast; irreversible ones wait for the Build position’s deliberation, however confident you feel.
Is the organization ready? The third gate asks about capacity rather than correctness. A change can be confirmed, reversible, and still wrong to launch, because the organization has nothing left to absorb it with. This is not a soft concern. Gartner’s research on change fatigue, reported in Harvard Business Review, found that employees’ willingness to support organizational change collapsed from 74 percent in 2016 to 43 percent in 2022, while the average number of planned changes an employee experienced in a year rose from two to ten. Attention is the scarcest resource a leader spends, and the third gate is where the spending is checked against the balance.
Notice what the gates never do: they never discard a finding. Fail the first and the finding goes to Watch, with a note about what would confirm it. Fail the second and the move goes to Build, where irreversible work gets the deliberation it deserves. Fail the third and it goes to Watch with a capacity trigger instead of an evidence trigger. The gates are not a filter. They are a router that puts every finding where it can be trusted.
The First Ninety Days, in Three Moves
With the portfolio filled and the gates passed, the plan needs a shape in time. Ninety days is the traditional window, and the right one, not because change finishes in a quarter but because credibility is decided in one. It divides into three moves of roughly thirty days each.
Show the Portfolio
- Deliver the readout: what you heard, what the evidence shows, and where the leadership team aligned.
- Name the Protect list first, and credit the people behind it.
- Announce the few Act moves, each traced to its finding.
- Share the Watch list openly, with owners and review dates.
- Start the first reversible move within the week.
Deliver the First Proof
- Land the first visible Act moves, and say what changed.
- Credit the teams and voices that surfaced each finding.
- Begin the Build work quietly, with the people it affects.
- Reverse anything that is not working, plainly and early.
- Keep listening: the tour ended, the habit did not.
Set the Rhythm
- Complete the remaining Act moves and close the loop on each.
- Put the Build commitment on its long-term footing.
- Hold the first scheduled Watch review, in the open.
- Settle the operating cadence: how decisions, reviews, and priorities will run.
- End the transition on purpose: ordinary leadership begins.
The first move is the hinge of the whole system. The readout is the Learning Readout the cornerstone promised: the first executive presentation, and deliberately not a strategy reveal. Its structure is the portfolio itself: here is what I heard, here is what the evidence shows, here is what stays, here is where we will act first, and here is what I am still watching. Delivered that way, the organization hears its own words in the plan, and the changes that follow feel fair because their origins are visible.
The second and third moves keep the two promises the readout makes: that the first actions will land, and that the watching will happen. Nothing erodes a new leader’s standing faster than an announced move that quietly stalls, which is why the Act position was restricted to moves that could be finished, and why reversals happen in daylight. If a move proves wrong, undo it, say what was learned, and return the finding to Watch. Done without defensiveness, a public correction reads as evidence the system works.
Momentum does not come from the number of things moving. It comes from the ratio of things finished to things promised. Three moves that land, one commitment that visibly deepens, and a protected core that never wobbled will carry more of the organization than a dozen simultaneous initiatives, every one of them half done. Fewer, finished, and fair beats many, loud, and stalled.
A working document for the operating phase: a findings-to-actions page, the Action Portfolio canvas, the Three Gates applied to your candidate moves, the 90-Day Arc, and a board-summary one-pager. Employer-agnostic, and designed to be filled in from the work of the earlier phases.
Download the workbook (PDF) →Prefer just the canvas? The Action Portfolio Worksheet (PDF) holds the portfolio and the Three Gates on two pages, ready to fill in during a working session.
A Reflection Before You Commit
Before the readout goes on the calendar, a short reflection. For each statement, consider how true it is today, from 1 (not yet) to 5 (fully, and I could point to why).
| # | Statement |
|---|---|
| 1 | My findings and my candidate actions live on separate pages, and every action can name its finding. |
| 2 | I can say, specifically, what I will protect and who built it. |
| 3 | My Act position holds four moves or fewer, and each is confirmed, reversible, and visible. |
| 4 | Nothing irreversible is scheduled for the first ninety days without deliberate, slower treatment. |
| 5 | I have honestly weighed what else the organization is absorbing right now. |
| 6 | Every finding I am not acting on has an owner, a reason, and a review date. |
| 7 | The readout is structured around what I heard, not around what I intend. |
| 8 | I know which of my first moves I would reverse, and what reversing it would cost. |
Mostly 4s and 5s: the portfolio is ready, and the readout will land as the product of listening rather than the end of it. A mix of 2s and 3s: the shape is right but a position is thin, most often Protect, which leaders skip because it feels like inaction, or Watch, which feels like admitting incompleteness. Both impressions are wrong, and both positions are load-bearing. Mostly 1s and 2s: the plan is still a task list wearing a new name. Go back to the dividing line: findings on one page, actions on the other, and let the gates do the sorting.
Where to Start
- Write the two pages. Findings on one, candidate actions on the other. Strike every action that cannot name its finding, and note every finding that has nowhere to land yet.
- Fill Protect first. Name what works, trace who and what is behind it, and draft the sentences that will tell the organization it stays.
- Run the gates. Take every candidate move through confirmed, reversible, ready, in that order, and let the failures route themselves to Build or Watch.
- Cut Act to four or fewer. If more than four moves passed the gates, keep the ones with the strongest evidence and the shortest path to visible completion. The rest keep their gate results and wait.
- Draft the readout in the portfolio’s order. Heard, shown, protected, acting, watching. If a section feels thin, that is the section that needs another week, not a better slide.
The End of the Beginning
This article completes the operating cycle the cornerstone promised: Listen, Map, Align, Operate. One activity faced the organization, months ago, in the form of honest conversations. Everything since has been the discipline of taking those conversations seriously: mapping how the place really works, testing what you believed you learned, and now converting the tested picture into commitments the organization can watch you keep.
What follows day ninety is not another phase. It is leadership as usual, practiced by someone who now actually knows the organization they are leading, with an operating rhythm to maintain, a Build commitment to see through, and a Watch list that keeps the habit of curiosity alive. For marketing leaders, this is also the handoff point this series has pointed to from the start: the place where organizational intelligence goes to work inside an operating system for the function itself.
And the system is not finished. The Organizational Intelligence framework extends beyond the first cycle, into the practices that keep an organization understood as it changes: the standing habits of listening, the instruments that age well, and the questions that deserve to be asked long after the transition ends. Those publications are next.
Key Takeaways
- Findings are not actions. Separate them physically, trace every action to a tested finding, and let no finding vanish silently. The organization is watching for its own words in the plan.
- Hold the first plan as a portfolio. Protect, Act, Build, Watch: four positions with scarcity built in, in which leaving something alone and watching something openly are decisions of equal rank with acting.
- Guard the Act position with the Three Gates. Confirmed, reversible, ready, asked in order. Reversible moves may go fast; irreversible ones get deliberation no matter how confident the evidence feels.
- Announce what will not change. Protection releases the organization’s bracing, credits the people who built what works, and is the cheapest trust a new leader can buy.
- Momentum is finished work, not activity. Three moves that land and one commitment that deepens beat a dozen initiatives half done. By day ninety, end the transition on purpose.
Frequently Asked Questions
What is the Action Portfolio?
The Action Portfolio is the operating instrument of the Organizational Intelligence System’s final phase: the complete set of a new leader’s first commitments, held in four positions rather than on a task list. Protect names what already works and will not change. Act holds a small number of confirmed, reversible, visible first moves. Build holds the one or two foundational commitments that will take longer than a quarter. Watch holds the findings that deserve continued observation, each with an owner and a review date. Every tested finding lands in exactly one position, so nothing is rushed and nothing is lost.
How do you decide what to change first as a new executive?
Ask three questions of every candidate move, in order. Is it confirmed: did the finding behind it recur across unrelated conversations, survive the written record, and get recognized by the people closest to the work? Is it reversible: if the move proves wrong, can it be undone at reasonable cost? Is the organization ready: is there attention and capacity for it right now? Moves that pass all three gates become first actions. Confirmed but irreversible work goes to the Build position for slower deliberation, and unconfirmed findings go under open observation. The order matters because it puts evidence ahead of instinct and trust ahead of speed.
Why should a new leader announce what will not change?
Because the organization is waiting to find out. Around every new executive, people brace for change: work slows, decisions wait for the new direction, and the most capable people quietly weigh their options. Naming what is protected, publicly and early, releases that tension. It tells the people who built what works that it was seen and valued, and it concentrates attention on the few changes that are actually coming. Preservation is a decision, and announcing it is one of the least expensive trust-building moves available to a new leader.
Which findings should wait, and is waiting just delay?
Waiting is a decision when it has an owner, a reason, and a review date; it is delay when it has none of those. Findings belong under observation when they are real but not yet confirmed, when acting on them would be hard to reverse, or when the organization has no capacity left for another change this quarter. The Watch position makes the waiting explicit: each item is written down with what would confirm it and when it will be looked at again, and the list is shared in the open. Held that way, waiting is the same patience that built the understanding, still working.
How do the first ninety days actually unfold?
In three moves of roughly thirty days. The first month is for showing the portfolio: the readout in which the leader shares what they heard, what the evidence shows, what will be protected, where they will act first, and what they are still watching, with the first reversible move starting immediately after. The second month is for delivering the first visible results and crediting the people who surfaced the findings behind them. The third month is for putting the deeper build work on its long-term footing, holding the first scheduled review of the Watch list, and settling the operating rhythm, so that by day ninety the transition has ended and ordinary leadership has begun.
What happens if a first move turns out to be wrong?
If the gates were respected, the damage is small and the recovery is fast, because the move was reversible by design. Undo it in daylight, say plainly what was learned, and return the finding to observation. Handled without defensiveness, a public correction can strengthen credibility rather than spend it: the organization sees that decisions trace to evidence, that course corrections come quickly, and that being wrong is survivable. What damages trust is not a reversed decision. It is a leader defending a mistake because reversing it feels like weakness.
Research & Supporting Evidence
The Action Portfolio, the Three Gates, and the 90-Day Arc are original ERM Advisory concepts, part of the Organizational Intelligence System. The transition and change research cited above is drawn from the primary sources below.
- Michael D. Watkins, The First 90 Days, Harvard Business Review Press (updated edition, 2013). Surveys of more than two hundred company presidents and CEOs put the break-even point of a typical newly hired or promoted leader at roughly 6.2 months; structured transition plans shorten it.
- Jeff Bezos, 2015 Letter to Shareholders, Amazon.com, Inc. (2016). The distinction between Type 1 decisions (one-way doors: consequential, nearly irreversible, deserving slow deliberation) and Type 2 decisions (two-way doors: reversible, best made quickly), and the observation that organizations err by treating one type like the other.
- John P. Kotter, “Leading Change: Why Transformation Efforts Fail,” Harvard Business Review (1995; republished 2007). From a study of more than one hundred companies: successful change efforts systematically plan and create visible short-term wins rather than hoping for them, and declaring victory too soon kills momentum.
- “How to Beat the Transformation Odds,” McKinsey & Company (2015). In a global survey, only 26 percent of executives said the transformations they knew best had succeeded at both improving performance and sustaining the improvement over time.
- Jessica Knight & Cian Ó Mórdha (Gartner), “Employees Are Losing Patience with Change Initiatives,” Harvard Business Review (2023). Gartner survey research: employees’ willingness to support enterprise change fell from 74 percent in 2016 to 43 percent in 2022, while the planned changes the average employee experienced in a year rose from two to ten.
Conclusion: Lead From What You Learned
The promise this series opened with was modest and demanding at once: understand the business before you change it. Keeping it took a quarter of disciplined work, and the temptation now is to treat action as the reward for patience, the moment the discipline can finally relax. It is the opposite. The first plan is where the discipline shows, or does not.
Held as a portfolio, the plan makes visible what three months of listening should have made true: this leader changes what the evidence supports, protects what deserves protecting, takes irreversible steps slowly, and says out loud what they are still unsure of. An organization that watches its new executive do those four things for ninety days does not need to be persuaded to follow. It has seen how it will be led, and it has seen its own fingerprints on the result. That is what organizational intelligence is for: not a wiser observer, but a leader whose first commitments hold, because they were built on what the organization already knew was true.