I do fractional operator work. That means, by definition, every engagement ends with a handoff. Over the last few years I have watched what makes the handoff hold and what makes it fail. The single biggest predictor is not who the successor is. It is whether the operating system was ever explicitly there to transfer.
Most succession plans focus on the person. The right ones focus on the machine. Here is what I do differently now.
Why naming the successor is the easy half
Every organization I have worked with can name a successor. Sometimes it takes a week of conversation, sometimes it is obvious, but the naming happens. And then most orgs treat the succession as complete once the name is on the org chart.
The naming is the easy half because it does not require examining the operation. It requires examining the people. It is uncomfortable, but it is not diagnostic. The hard half is asking: what is the outgoing operator actually doing, day to day, week to week, month to month, and is that work transferable at all?
The answer, in most operations, is that a meaningful amount of the outgoing operator's real work is not transferable in its current form. It lives in their head. Their inbox. Their phone. Their history with vendors and landlords and key employees. If they walked out today, the operation would experience a set of quiet failures over the next six months that nobody would trace back to the handoff.
Operating system succession is the discipline of making sure that does not happen.
Fig. 1 · The 90-day operating system succession.
What actually gets transferred
An operating system, in the sense I use the term, has six components. Any real succession has to transfer all six explicitly.
1. The standing meetings
Every recurring meeting the operator runs or attends: weekly one-on-ones with direct reports, weekly P&L reviews, monthly regional operating reviews, quarterly business reviews, vendor reviews, board updates. Each meeting has an agenda, an owner, a cadence, and a purpose. All of it goes on paper.
2. The dashboards and reports
Every number the operator personally watches or is asked about. Where the data comes from. When it refreshes. What "good" looks like. What triggers escalation. If the outgoing operator opens a dashboard every Monday morning, the successor needs to know why, how to open it, and how to read it.
3. The operating rhythm
The choreography of the week and month that pulls everything together. Monday flash review at 8am. Tuesday store visit. Wednesday regional call. Thursday P&L close prep. Friday planning for next week. This is the meta-schedule that most operators keep in their head. Written down, it becomes a machine.
4. The accountable metrics
The four or five numbers the operator is personally accountable for. What target. What definition. Where they land today. What levers move them. This is often the last thing to get transferred because it feels obvious to the outgoing operator, and it is often the thing the successor gets wrong first.
5. The recurring decisions
The choices the operator makes on a regular cadence: pricing changes, hiring approvals, capital requests, comp policies, marketing spend. Each has criteria, a decision framework, and a history of prior decisions that inform the next one. Successors who inherit only the decisions without the framework end up either reinventing bad rules or being paralyzed by the first hard call.
6. The relationships
The people the operator knows and calls. Key vendors. Landlords. Long-tenured employees. Retail partners. Board members. Regulators. Each of these relationships has history and unwritten context that only transfers with direct introduction.
Everything the outgoing operator knows that only they know is a liability, not an asset. The job of the last 90 days is turning private knowledge into shared operating system.
The 30-30-30 succession sequence
Days 1 to 30: The successor observes
The successor sits inside the running rhythm for a full month. Every meeting the outgoing operator attends, the successor is in the room. Every dashboard the outgoing operator opens, the successor opens too. Every phone call, every field visit, every decision. They are silent. They are watching.
Nothing is being handed off yet. This phase is dedicated to understanding the machine as it is running. If the successor tries to change anything in these 30 days, they are guessing. They cannot see the whole yet.
At the end of the 30 days, the successor should be able to describe the operating rhythm from memory. Every meeting, every dashboard, every metric, every relationship. If they cannot, the observation phase is not done and phase two waits.
Days 31 to 60: The successor drives, the outgoing operator observes
One artifact per week transfers. Week one: the weekly P&L review. The successor runs it. The outgoing operator sits in the room and does not speak unless asked. After the meeting, they debrief privately.
Week two: the general manager one-on-ones. Week three: the monthly regional review. Week four: the vendor review or the landlord conversation or whatever cadence lands next.
By the end of day 60, the successor has run every recurring artifact of the operating system at least once with the outgoing operator watching. The debriefs after each one are where the tacit knowledge finally transfers. "I would have pushed harder on the labor number." "I would have opened with the wins, not the misses." "I never let that vendor set the agenda."
Days 61 to 90: The successor owns, the outgoing operator coaches
The successor is now running the operation. They open the dashboards. They chair the meetings. They make the decisions. The outgoing operator is available for judgment calls and hands off the last few relationships that require personal introduction. But they are no longer in the room.
This phase is where confidence gets built or breaks. If the operating system was transferred well in the first 60 days, the successor will get 80 percent of the calls right, ask the outgoing operator about 15 percent, and get the last 5 percent wrong in ways that do not damage the operation. If the system was not transferred well, the successor will freeze on the first hard call and start pinging the outgoing operator daily.
The two-test check for whether the operating system is ready
Fig. 2 · Two tests that tell you whether the handoff actually worked.
Test one: can the successor run last week's operating rhythm without asking a question? Using only what is written down. Not what they remember from the debrief. Not what they can find in the outgoing operator's inbox. What is documented, published, and part of the operating system.
If they cannot, something in the operating system is still tacit. Find it. Get it out. Then test again.
Test two: does the P&L stay steady in the 90 days after the handoff? Not perfect. Steady. Labor within a point of the target trend. Food cost within a point. Comps and voids within normal range. Sales tracking to forecast.
If the P&L drifts in the first 90 days after the outgoing operator leaves, the transition was a communication exercise, not a system transfer. The good news is you can usually fix it if the outgoing operator is still available. Bring them back for a two-week diagnostic. Find what did not transfer. Transfer it. Then let them go again.
What I have watched fail
The 30-day handoff
Most orgs try to run succession in 30 days because that is what the calendar allows. Under 90 days, the successor is guessing. They have not seen a full monthly cycle. They have not run a P&L close. They have not sat through a full vendor cycle. They walk in on day 31 with confidence and no context, and the operation drifts.
The "written playbook" substitute
Some outgoing operators try to substitute a document for the 90-day handoff. They write a beautiful 40-page playbook and hand it to the successor with a cup of coffee. Six weeks later, the playbook has been read once and the successor is running the operation the way they think it should be run, based on their own history, not the operating system the outgoing operator built.
The playbook matters. It is not sufficient. Running the machine together is what actually transfers the operating system.
The invisible relationships
The failure I have seen most often is on the relationship transfer. A vendor who has been extending payment terms as a favor to the outgoing operator. A landlord who has been quietly holding rent flat because of a personal relationship. A key long-tenured employee who stays because the outgoing operator asked them to. None of these are on the org chart. All of them will fail without deliberate transfer.
The fix is a relationship inventory in the first 30 days. Every key relationship, named, with a history, with a state, and a plan for how it gets introduced.
What holds after the outgoing operator leaves
The gains that hold, past the succession, are the ones written into the operating system. The gains that fade are the ones held in the outgoing operator's head. That distinction is the whole point of documenting the operating rhythm while you are running it, not while you are leaving.
I write down the operating rhythm from the first month of any engagement, not the last month. Not because I plan to leave. Because writing it down forces me to be clear about what the operation actually is, and clarity is what makes it repeatable. If I never leave, I have a cleaner system to run. If I do leave, the operation survives the leaving.
The point
An operator succession is a system transfer, not a communication exercise. Ninety days minimum. Thirty of observation, thirty of driving with backup, thirty of coaching from the side. Six components of the operating system transferred explicitly: meetings, dashboards, rhythm, metrics, decisions, relationships.
The test is not "does the successor understand the plan." The test is "can they run last week's rhythm on Monday without asking a question, and does the P&L hold in the 90 days after that."
Cadence beats charisma. That is true when you are building the operating system, and it is even more true when you are handing it over. The best succession I have ever run was the one where the successor said, at day 60, "I already know what to do here. The machine is telling me." That is what an operating system succession sounds like when it works.