The Monthly Leadership Session: What We Actually Discuss During Run

A CEO I work with sat down for his third monthly leadership session last week. Twenty minutes in, he stopped me mid-sentence. "I just realized something. You haven't actually pulled out a checklist. Everything we're talking about came out of my own system this morning."

He was right. The agenda that day was a decline in one automation's approval rate, an uptick in his Away-From-Desk Autonomy number, and a question his COO had queued up about opening a second service line now that her team had real headroom. None of it came from a template I brought. All of it came from what the AIOS reported over the previous 30 days.

That small realization is, honestly, the whole point of Run.

Why it is not a generic advisory retainer

Most consulting retainers are some version of "call me when you need me." A standing line of credit on senior attention, billed monthly whether you use it or not. The client drives the agenda. The consultant shows up and reacts.

Run is built the other way around. The monthly leadership session is not a reactive call. It is a recurring, structured session whose agenda comes out of the telemetry the system itself is producing. Where Layer 3 is getting stuck. Where Layer 4 is approving faster than the trust envelope suggests it should. Where Layer 5 builds are slipping. Where headroom is opening on the team that wasn't there a quarter ago.

That distinction matters commercially and operationally. A reactive retainer rewards consultants who are good at being available. A telemetry-driven session rewards the ones who are good at interpreting what a working system is telling you. It is closer in spirit to what HBR has written about strategy execution than it is to traditional advisory. Execution, not advice, is the center of gravity.

We keep the session to 60-90 minutes. Monthly cadence. No skipping. If the telemetry is quiet, we use the time for the headroom conversation, which I will get to.

The standing agenda

The session has six segments. The shape is standing. The content inside each is different every month because it is set by what the system did.

1. State of the system (10 min). We walk the Layer 1-5 snapshot for the month. What Layer 1 captured that it wasn't capturing before. What Layer 2 cleaned up or surfaced. What Layer 3 produced in briefs and where its confidence was wobbly. What Layer 4 approved, overruled, or escalated. What Layer 5 shipped, broke, or retired. We also flag trust events. A trust gain looks like "Layer 4 ran this approval 42 times with zero overrides, we can loosen the gate." A trust loss looks like "the briefing agent missed a material pattern twice, we are tightening its scope."

2. Away-From-Desk Autonomy review (10 min). How many hours this month the CEO could fully step out of the operation without things degrading. We track it monthly because it moves. A new automation ships and it jumps. A key hire leaves and it drops. We compare this month against last month and against the rolling three-month average. It is a moving target, and the movement is the insight.

3. Task Automation percentage trend (10 min). The headline KPI. The target band is 60-70 percent, and I have written at length about why we cap it there rather than chasing higher numbers in the 60-70 percent automation target. In this segment we look at where the number currently sits, whether it moved up or down, and whether any automations have become pruning candidates. Pruning is the unsexy part of Run. Roughly once a quarter, something we built earlier in the engagement is no longer worth keeping alive, and we decommission it on purpose.

4. Where headroom is opening (20 min). This is the longest segment and it is the part most leaders have never actually done before. When the system is working, the team has bandwidth it did not used to have. That bandwidth is an asset and most firms waste it, either by filling it with busywork or by hiring ahead of the demand curve. The leadership session forces the question the other direction. What new capability does this team have headroom for now. What new service line, new market, new build, new customer segment becomes possible in this configuration that wasn't possible six months ago. I will go deeper on this segment in the next section, because it is where the session gets genuinely interesting.

5. Cross-layer upgrades (15 min). The Run work we are committing to for the month ahead. Which new automations are entering Build. Which existing ones are being retuned. Which ones are being decommissioned. Which handoffs between layers need a tighter contract. This segment produces the Build backlog for the next 30 days, which is why the entire Run phase still has an active Build pipeline rather than just a maintenance function.

6. The standing closing question (5 min). I close every session the same way. I ask whether the system is paying for itself. If the answer has been yes for 12 or more consecutive months, we start talking about graduation. I have written about what graduation actually means in when an engagement ends: graduation, not churn. The short version is that the goal of a good Run relationship is to eventually not need us.

Where headroom conversations get interesting

Of those six segments, the headroom conversation is the one most CEOs are least prepared for the first time it comes up.

The pattern is consistent. Month two or three of Run, the Task Automation percentage clears 55. Away-From-Desk Autonomy clears 20 hours a month. The COO's inbox isn't the bottleneck anymore. And in the leadership session I ask, "what does your team have room to take on now that they didn't six months ago." The first time I ask, the CEO usually pauses, because the question itself is unfamiliar. They have spent years optimizing for capacity constraints, and now the constraint is no longer the limiter.

The conversations that come out of that question are the highest-leverage ones (noun sense) we have during the whole engagement. New service lines that were tabled because staffing them felt impossible. Adjacent customer segments the ops team never had time to serve. Build projects on Layer 5 that were deprioritized because the team couldn't absorb them. Acquisitions. Partnerships. Sometimes a decision to slow down and take margin out of the system instead of growth, which is a legitimate answer.

This segment is also where the AIOS stops being a cost center in the CEO's mental model and starts being a growth enabler. The change shows up in how they talk. Early Run, they talk about what the system is doing for them. Mid to late Run, they start talking about what they can now do because of it. The revenue per employee number usually starts moving around the same time, which is not a coincidence. Good reading on how this kind of operating capacity translates into strategic moves lives in the MIT Sloan Management Review's AI coverage and McKinsey's organizational performance work.

Who should attend, and who shouldn't

I get this question on almost every engagement kickoff, so I am going to be direct.

The CEO attends. Every month. No exceptions. This is not a session where a chief of staff can sit in as proxy. The decisions about pruning, retuning, headroom, and graduation are all CEO-level calls.

The COO attends if the firm has one, which is most of the time in the $1M-$50M band. The COO is usually the person closest to the day-to-day operational reality the telemetry is describing, and their instinct on whether a reported pattern matches lived experience is load-bearing.

The CFO sometimes attends, typically around budget cycles, around pricing changes, or when the "is the system paying for itself" question is being answered with real numbers for a board meeting. Not every month.

Nobody else. I have learned this the hard way. The full leadership team does not attend. Function heads do not attend. The session diffuses the second it has more than three or four people in it, because the headroom conversation in particular requires the kind of candid acknowledgment of weaknesses and priorities that most leaders will not do in front of their direct reports. If a function head needs to be looped in on a specific decision from the session, the CEO handles that in a separate conversation.

One working session. Small room. Telemetry on the screen. That is the format.

The standing closing question

I always close with the same question. "Is the system paying for itself."

The answer is almost always yes by month four. By that point the Task Automation number is usually in band, Away-From-Desk Autonomy is a real figure we can hold up against a cost of the CEO's time, and there is usually a revenue or margin impact that's attributable to capacity that didn't exist before.

But I ask every month anyway. Because the day the answer is no, we need to know, and we need to do something about it. That might be pruning automations that aren't pulling weight. It might be retuning a layer that has drifted. It might be an honest conversation about whether the engagement should end, which is a conversation I am willing to have and most retainer-model consultancies aren't. The willingness to ask the question is what keeps the relationship honest. If you cannot tell me whether your AIOS is paying for itself, either the measurement is broken or the system is broken, and either way that is the highest priority for the month.

Twelve consecutive yes answers starts the graduation conversation. I like that outcome. It means the firm's operating leadership, not the consultant, now owns the AIOS, and the 24-month window for getting there has done its job.

How this session shapes the Build backlog for the month

Something worth making explicit, because it is often misunderstood. The leadership session is not a pure review meeting. It produces work.

Every session closes with a short written set of Run decisions for the month. New automations committed to Build. Existing ones scheduled for retune or decommission. Layer contracts to tighten. Headroom bets the leadership wants the firm to pursue, which sometimes means new capabilities the AIOS needs to support. That document is short, maybe a page, and it becomes the Build backlog for the 30 days ahead.

Which means the monthly leadership session is not an extra on top of the engagement. It is the governance mechanism for the engagement. The telemetry feeds the session. The session produces the backlog. The backlog drives Build for the month. Build ships, which feeds the next month's telemetry. That loop is the actual thing we are selling in Run. Not hours. Not availability. The loop.

If you are exploring whether your firm is in a shape where this loop would work, start with the Fit Check. It will tell you honestly whether you are ready, and if you are not, what to fix first.

-Ed