I have been installed as CEO of a company I did not build four times. Three of those times, the same playbook worked exactly as designed: audit the pricing, the acquisition channels, and the win-loss data, present the findings, and get to work on the fix. The fourth time, at Sendspark, I ran that playbook in the first month and nearly cost myself the team before I had earned any right to lead it. The data in the audit was accurate. The timing was wrong, and the difference between those two things is the subject of this essay.
Why the 100-day plan works when someone else already made the deal
A PE-installed CEO’s 100-day plan works because the authority behind it is negotiated before the CEO ever walks in. The board has already agreed on a thesis, so an audit of pricing, channels, and win-loss data lands as confirmation of a decision the team already expects, not as a challenge to it.
At FIVE CRM, where I became Senior General Manager for Xenon Partners in February 2022, the board had already concluded the company needed a pricing and channel reset before I had read a single account file. My job for the first two weeks was to put real numbers behind what they suspected: which segments were underpriced, which acquisition channels were burning budget without producing customers who stuck around. I sent the audit to the leadership team in week three. Nobody in that room was surprised by the conclusions. They were relieved someone had finally written them down in one place.
The same shape held at Teamgate Sales CRM, where I took over that April, and again at Appsembler starting in July 2023. Different products, same starting condition. A deal had just closed, a board had a thesis, and the team was waiting to find out which parts of that thesis the data confirmed. In every one of those rooms, the audit was not received as an opinion. It was received as a verdict on a decision that had already been made above the team’s head, which is exactly why it worked as an opening move. (The full operating history behind all four roles, including the two years at Xenon Partners, is on my about page.)
The playbook depends on that condition existing before you arrive: an authority granted by people who already control the room. Sendspark did not give me that condition, and I did not notice until the damage was already done.
What’s different when the founders already left
At Sendspark, there was no deal thesis to confirm and no board that had pre-approved a mandate for change. Both founders had already left before I became CEO in December 2025, which meant the team’s opening question was not “what will the new owner change.” It was “can we trust the next person at all.”
That distinction changes what the first artifact needs to be. In a PE turnaround, the team knows a transaction happened and is waiting to learn the terms. At Sendspark, nobody had voted for an acquisition, because there wasn’t one. There was no data room, no formal handover memo, no board explaining the thesis to the people who would have to live with it. There was a leadership gap, and then there was me, standing in it with a folder of diagnostic questions I hadn’t earned the standing to ask yet.
A team that has just watched the people who built the company walk out is not looking for a diagnosis of what’s broken. It’s looking for a reason to believe the next person understands what they already know. That reason has to be earned before any audit gets read charitably, and I skipped straight to the audit. This is a different failure mode than the founder-insertion problem I’ve written about before, which is about a founder coming back into decisions after already stepping back. This one is about what happens when there is no founder left to come back at all. Both belong to the harder half of founder lessons that survive a change in who’s actually running the place: the operating discipline that works when you’re brought in to execute someone else’s decision does not automatically transfer to a room where the decision to trust you hasn’t been made yet.
The specific thing I got wrong in the first month
I opened at Sendspark with the exact file structure that had worked three times before: a written audit of the sales motion, the gaps in the product roadmap, and friction in the customer onboarding flow, sent to my direct reports inside the first two weeks. I had not yet spent real hours listening to how the team itself described the business, and the file read as a verdict from someone who hadn’t earned the standing to render one.
I want to be specific about what went wrong, because “I moved too fast” isn’t a lesson anyone can use. The audit itself was not inaccurate. The gaps I flagged were real gaps, later confirmed by the same people who first pushed back on the file. The problem was that I presented them as if I had discovered something the team had missed, when in most cases the team already knew about every one of them and had reasons, some good, some outdated, for why nothing had changed yet. One of my direct reports asked me, in the meeting where I walked through the findings, how I could already know what needed fixing after twelve working days in the building. I did not have a good answer, because there wasn’t one to give.
The mistake was not the content of the audit. It was sequencing: I had substituted a document for the months of context that make a document trustworthy. In a PE turnaround, that context already exists in the deal thesis. At Sendspark, the context didn’t exist yet, and I had tried to skip the step where I earned it.
The consequences showed up in small, specific ways over the following two weeks rather than in one dramatic confrontation. Meetings that had been candid got shorter and more careful. A couple of the changes I’d proposed as low-priority notes got treated as decisions I’d already made, because the team had no way yet of knowing which parts of the file were firm and which were open questions. I found out later that at least one person had quietly started updating a resume, not because the audit was wrong about anything, but because it read as the opening move of someone who intended to restructure first and understand the business second. That is the actual cost of getting the sequence backward. It is not that the data gets rejected. It is that the team starts protecting itself against a leader they haven’t decided to trust yet, and every subsequent interaction gets filtered through that caution instead of taken at face value.
The difference between installed authority and earned authority
An installed PE CEO’s authority is granted by the cap table before day one. A founder-succession CEO has no equivalent grant and has to earn the same authority through the team’s own judgment about whether the new leader understands what actually happened before they arrived. Getting that sequence backward, running the audit before the listening, is the highest-risk mistake available in the first month of either job.
| PE-installed turnaround | Founder-succession CEO | |
|---|---|---|
| Mandate source | Board-approved deal thesis | The team’s own judgment, earned over time |
| Day-one authority | Pre-negotiated before arrival | None; has to be built |
| Right first artifact | Diagnostic audit | A listening file, built from the team’s own account |
| Biggest risk if the sequence is reversed | Slower execution | A trust deficit that outlasts any later correction |
The table understates how asymmetric the risk actually is. A PE-installed CEO who audits too slowly loses a few weeks of runway toward a target the board already agreed to wait for. A founder successor who audits too fast can lose the benefit of the doubt permanently, because the team’s first data point about the new leader becomes “arrived with conclusions” instead of “arrived with questions.” You do not get many chances to reset that first impression. I’ve written before about the AI agent version of this same asymmetry: the cost of moving fast on a wrong assumption is rarely symmetric with the cost of moving slow to get the assumption right, and I had the data on that point long before I ignored my own advice in December.
What I changed after getting it wrong
I pulled the written audit back for three weeks and replaced it with unscripted conversations, one with almost everyone on the team, asking what already worked before I asked what I would change. The output was a short file describing what the team itself said was working, written mostly in their own language, which I read back to them before I wrote a single recommendation.
The conversations were not structured interviews. I asked open questions about what they were proud of, what they thought outsiders misunderstood about the company, and what they’d protect if a new leader tried to change it. I wrote down patterns, not individual complaints, and I did not attribute anything to a specific person when I circulated the notes. The “what already works” file that came out of it was not a strategy document. It was closer to a mirror: a description of the business that the team recognized as accurate, in some cases more accurate than my own early read had been.
The “what I’d change” file came weeks later, and it referenced the listening file explicitly wherever it proposed something different, so people could see their own account reflected in the reasoning behind any change. That single structural choice, citing the team’s own words back to them as the basis for a proposal instead of presenting the proposal as a conclusion I’d already reached alone, was the difference between a room that argued with me and a room that revised the plan with me.
The other thing that changed was pace. At FIVE CRM and the other Xenon Partners roles, moving fast in the first month was the entire point of being installed. At Sendspark, I slowed the calendar down on purpose. I pushed the first all-hands past where I’d normally schedule it, because I did not want to stand in front of the team with a plan before I had spent enough hours in their world to have a plan worth defending. Some of that slowness cost real time against whatever a board would have measured if there had been one. There wasn’t, so the actual cost was smaller than my instincts told me it would be, and the trust it bought back was worth more than the weeks it took.
What the research on new leaders actually says
Between one-third and one-half of new CEOs fail within their first 18 months, whether they were promoted internally or hired from outside, according to Harvard Business Review research by Dan Ciampa. The failure mode most often traces back to decisions made in the earliest weeks, which is exactly the window where I ran the wrong artifact first at Sendspark.
The scale of the problem isn’t limited to CEOs. Joint research from Genesis Advisers, Harvard Business Review, and IMD, summarized on IMD’s “First 90 Days” program page, surveyed more than 1,300 senior HR leaders and found that close to 90% agree transitions into new roles are the hardest stretch of a leader’s career, while roughly three-quarters agree that success or failure in the first few months predicts success or failure in the job overall. The first weeks are not a warmup. They are closer to the whole test, compressed.
McKinsey’s own playbook for newly installed private equity portfolio-company CEOs frames the standard diagnostic test as whether a PE firm would buy the business today and keep its existing talent. That’s a good test when a board is actually deciding whether to buy the business. It is exactly the wrong question to run silently against a team that was never for sale, and running it anyway is a fair description of what my first audit did without my noticing. The Harvard Law School Forum on Corporate Governance has made a related argument at the board level: the popular fixation on a new CEO’s first 100 days oversells what actually predicts whether the tenure holds up two or three years out, and a plan built to impress a board in the first quarter is not the same plan that earns a team’s trust in the first quarter.
None of that research is about founder succession specifically, which is itself worth noting. The literature on new-leader failure is built almost entirely on situations that already resemble a PE turnaround: a board, a mandate, a formal process. The founder-succession case, arriving after the people who built the company have already left and nobody voted on the arrangement, is the harder version of a problem that’s already this well documented as failing half the time. That should lower everyone’s confidence in the standard playbook’s portability, mine included.
What this means if you’re inheriting a team you didn’t build
Build the trust file before the data file. If nobody in the room chose you, the first weeks are not for proving you can run an audit. They are for proving you already understand what the team knows that you don’t yet, and the order in which you produce evidence of that understanding is not a stylistic choice. It’s the whole decision.
The audit itself is still worth doing. Nothing about this argument is against rigor, diagnosis, or the discipline of finding out where the pricing is wrong or the roadmap has drifted. The claim is narrower: that work has to come second, after the team has heard its own account of the business reflected back accurately by the person now leading it. Skip that step, and the most accurate audit in the world reads as evidence you weren’t listening, which is a worse first impression than being slightly wrong about the pricing.
I keep coming back to how avoidable this was. I had already written, in an earlier essay on this site, that the cost of moving fast on an unverified assumption is rarely symmetric with the cost of moving slow enough to check it. I applied that logic to an AI rollout and ignored it walking into my own team’s first month, which is its own small lesson about how easily operating wisdom stays theoretical until it costs you something real. More essays on exactly this kind of gap between what I know and what I actually do under pressure live at dearmer.com.au.
If you inherited your current team instead of building it, what was the first document you sent them, and knowing what you know now, would you send that same one again?
Frequently asked questions
Why does a 100-day plan work for a PE-installed CEO but not a founder successor?
A PE-installed CEO's authority is pre-negotiated by the board before arrival, so a diagnostic audit confirms a decision the team already expects. A founder successor has no such mandate. The team never voted for the change, so the same audit reads as a stranger's verdict rather than a confirmation, and it has to be earned first.
What is the biggest mistake a new CEO makes replacing departed founders?
Sending a written diagnosis of what needs to change before spending real time listening to how the team itself describes the business. It signals competence in a PE turnaround. In a founder succession, it reads as indifference to people who mainly want to know whether the next leader can be trusted at all.
How common is it for new CEOs to fail in their first 18 months?
Between one-third and one-half of new CEOs fail within their first 18 months, according to Harvard Business Review research by Dan Ciampa, and this holds whether the CEO was promoted internally or hired from outside. The failure mode most often traces back to decisions made in the earliest weeks.
What should a new leader do differently when inheriting a team instead of building one?
Build a listening file before a change file. Ask the team what already works, in their own words, before proposing anything different. That sequencing signals the leader is trying to understand the business before rewriting it, which is the trust a founder successor has to earn that a PE-installed CEO gets for free.
Does the standard PE 100-day plan still have value outside of a PE deal?
Yes, but only in the right order. The diagnostic instinct, mapping what's working against what isn't, is still useful. It just cannot be the first artifact a founder successor produces. It works once the team has already told its own version of the story and recognizes itself in the analysis that follows.
How do you build trust with a team that did not choose you as their leader?
Spend the first weeks asking questions instead of presenting conclusions, and write down what the team says is already working before proposing what should change. Trust comes from a team recognizing its own account reflected back accurately, not from a leader demonstrating how quickly they diagnosed the business.
Sources & references
- Harvard Business Review — After the Handshake · Dan Ciampa's HBR research documenting that between one-third and one-half of new CEOs fail within their first 18 months, regardless of whether they were promoted internally or hired externally.
- IMD — The First 90 Days · Summary of the Genesis Advisers, Harvard Business Review, and IMD joint research surveying more than 1,300 senior HR leaders on why leadership transitions are the hardest stretch of a career.
- McKinsey — A Playbook for Newly Minted Private Equity Portfolio-Company CEOs · McKinsey's framing of the standard PE-installed CEO diagnostic, including the test of whether a PE firm would buy the business and keep its talent.
- Harvard Law School Forum on Corporate Governance — Beyond the First 100 Days Rhetoric · Argues that the popular fixation on a new CEO's first 100 days oversells what actually predicts long-term success in a leadership transition.