Designing the Handover to a Hired CEO: The Plan That Has to Exist Before They Start

Most handovers to a hired CEO are not designed. They're scheduled. A start date goes on the calendar, an onboarding checklist gets assigned, the founder gives a few weeks of overlap out of courtesy, and everyone calls that the transition. It isn't one. It's a start date with a goodbye attached.

A real handover is a transfer of authority and judgment, not paperwork — and like any transfer of something valuable, it has to be designed before it happens, not narrated afterward.

The plan starts too late, almost every time

By the time most firms think seriously about the handover, the CEO has already been hired, sometimes already started. That's backwards. The handover should be designed before the CEO search even begins, because the shape of the plan determines the kind of CEO you actually need — someone who can operate for two years in a founder's shadow before taking the wheel looks different on paper than someone who's expected to run the company from day one.

This is the same failure covered from the other direction in the founder-dependency assessment: skip it, and you end up applying a short, generic onboarding plan to what is actually a multi-year transfer of judgment nobody wrote down, because nobody stopped to ask how much of the company still lives in one person's head.

What the plan actually has to specify

A handover plan that survives contact with reality answers four questions before the CEO's first day, not after:

• Timeline, in years or quarters, not weeks — set by how founder-dependent the company actually is. A company with a strong second layer of management needs a short, clean handover. A company that is the founder needs an overlap measured in years, with the CEO's authority expanding on a defined schedule rather than all at once.
• Decision rights at each stage — what the CEO can decide alone on day one, what still routes through the founder, and exactly what has to happen for a decision to move from the second column to the first. Without this written down, the boundary gets negotiated informally, every time, which is slower and more political than either person wants.
• A trust protocol for the relationships that matter — which customers, which employees, which partners hear directly from the founder that this is real, and in what order. Authority that isn't personally and visibly transferred doesn't transfer; it just goes quiet until someone tests it.
• The founder's own exit conditions — what specifically has to be true (not just how much time has to pass) before the founder's role tapers from present, to consulted, to gone. Vague timelines produce vague endings, and vague endings are where founders check out early because nobody ever told them when they were allowed to.

The trust protocol is the part everyone skips

It's tempting to treat the handover as a knowledge-transfer problem — get the CEO briefed, get the documents written, done. But the people actually watching the handover aren't reading the documents. They're watching whether the founder personally, visibly, is putting their own authority behind the new CEO, especially the people who were loyal to the founder rather than to the company. A founder who says the right things in an all-hands and then keeps making the real decisions privately isn't handing anything over — they're running a shadow company with an org chart nobody follows.

The trust protocol is specific and sequenced on purpose: which customer calls the founder makes jointly with the CEO before stepping back from them entirely, which internal decisions get publicly redirected to the CEO even when it would be faster for the founder to just answer, which meetings the founder deliberately stops attending so the room learns to function without them. Every one of those moments is a signal, and the org reads the signals far more carefully than it reads the announcement.

The Board's Role: Co-Author, Not Just Approver

Most boards see the handover plan for the first time as a line in a board deck — approved in one meeting, then revisited only if something visibly breaks. That's backwards. A plan the board didn't help build is a plan the board doesn't understand well enough to judge when it's quietly failing. The board — or, on a smaller deal, the deal team and investment committee acting as one — should be in the room with the founder and the incoming CEO while the plan gets built: the timeline, the decision rights, the trust protocol, the exit conditions. Not receiving a summary of decisions already made.

Being a co-author carries concrete obligations, not just a vote of confidence at signing. The board should agree the founder-dependency assessment before the CEO search even starts, so everyone is aligned on how long the handover genuinely needs to run. It should sign off on the specific milestones that define the founder's exit conditions, so "done" isn't declared unilaterally by whoever is most eager to call it finished. And it should own the escalation mechanism in advance — because a mechanism invented after trust has already broken down is usually invented badly, under pressure, by whoever is angriest in the room.

How progress gets reported matters as much as what gets reported. An update that only covers financial performance tells the board nothing about whether the actual transfer of judgment is happening — the numbers can look healthy for months while the founder has already quietly checked out. Progress should be reported against the plan's own milestones: which relationships have actually been handed off, which decisions the CEO has made independently and how they turned out, whether the founder's exit conditions are on schedule. And it should include the founder's own account, not only the CEO's — a handover reported exclusively through the new CEO risks becoming a performance review the CEO writes about themselves.

This is also where the uncomfortable question has to live: what happens if the CEO turns out to be the wrong hire, or the relationship with the founder breaks down beyond repair? Most handover plans are written as if that possibility doesn't exist — which is exactly why, when it happens, nobody has the standing or the appetite to act on it quickly. The board should agree in advance who has the authority to make that call, what evidence would trigger it — a pattern, not one bad quarter — and what the fallback actually looks like: an extended founder role, an interim operator, a second search. A board that has already agreed on this in the calm of planning makes a far better decision than one making it for the first time in a crisis, with sunk cost and awkwardness both pulling toward "give it more time" long after more time has stopped helping.

Tapering beats a cliff, every time

The instinct is to structure the handover as a clean cutover — founder out, CEO in, effective a specific date. It's clean on the cap table and almost never clean in practice. The founder-dependent knowledge doesn't transfer on a fixed schedule; it transfers when a live decision forces it out, and live decisions don't wait politely for the handover to finish.

A better shape tapers in stages: a period where the founder still leads and the CEO builds real authority alongside them, a period where the CEO leads and the founder is reachable for the decisions that still need the old pattern recognition, and only then a clean exit — by which point the CEO has actually accumulated standing instead of just inheriting a title. The exact lengths vary by company. The sequence doesn't.

What this looks like once the CEO has started

A handover plan doesn't end when the CEO walks in the door — it's tested, in real time, across the entire first year under the new CEO, which is where you actually find out whether the plan held or was just a document everyone signed and quietly ignored.

Designing that plan well before the CEO's start date — the timeline, the decision rights, the trust protocol, the founder's own exit conditions — is easier with someone sitting between the deal team, the founder, and the incoming CEO, translating in all three directions until the handover is a sequence the organization can actually follow, not an announcement it has to take on faith.