Leadership and the operating system · Answered by Golden Section from more than 400 B2B software companies observed
Judge the executive against the seat as it will be in the next 12 to 18 months, not against their history or your loyalty. The signal is a seat whose written outcomes keep missing after you have ruled out other causes, such as unclear goals, missing resources or the wrong incentives, combined with a next stage that needs work they have never done and are not learning fast enough to do. Separate scope from effort. A performance gap goes on a dated ladder that ends in a decision; a scope gap needs an honest conversation now, and our default is a clean exit, because no improvement plan closes it. When you decide to end it, act within weeks, with counsel, clearly and with dignity, and have the interim owner named before the conversation.
Root-cause the misses first, then decide whether the gap is effort or scope. Effort goes on a written ladder with an end date and a decision at the end; scope gets a direct conversation that defaults to a clean exit, made before the company's next stage exposes it.
| Metric | Value | What it means | Source |
|---|---|---|---|
| Consequence ladder timing | 30-day check, then a plan of up to 60 days ending in a decision | conversation, written expectations, improvement plan, decisionapplies to executives as written; dishonesty skips the ladder; counsel reviews each plan before use | Golden Section playbookUnderperformer Consequence Ladder |
| Outcomes per seat | three to five | measurable outcomes each seat on the accountability chart ownsthe standard an executive is judged against | Golden Section playbookExecution Operating System |
| Typical hiring success rate | roughly one in four | share of hires that work out without a structured processwhy the replacement search should run through a scorecard, not a network favor | Golden Section playbookHiring A Players (Topgrading) |
Founders wait too long on executives for understandable reasons: the person helped build the company, the founder is not sure the problem is the person, and there is no stated process, so every case feels personal. Meanwhile the seat underdelivers, the team routes around it, and the strong performers notice first. By the time the founder acts, the team has already concluded what the standard really is.
Two tools make the call less personal. The accountability chart turns the question into one about a seat: what must this function deliver next year, and can this person deliver it? The consequence ladder handles effort gaps with dated steps and a decision at the end. Scope gaps are different. An executive who ran a 6-person team well may not be able to build a 30-person one, and no plan fixes that. Say so early, default to a clean exit and offer a different seat only in the rare case it is real, and hire the replacement through Topgrading. Read the employee agreement and any equity terms with counsel before the conversation, and use the board for air cover on the decision.
This company is invented. A VP of Customer Success joined at $1.5M in annual revenue and built onboarding from nothing. At $6M, gross retention has slipped two quarters running and the plan calls for a team of 15 across three segments. The CEO root-causes the churn: the playbook is sound, but nobody is managing the managers, and the VP still runs most escalations personally. It is a scope gap, not an effort gap. The CEO tells the VP directly, agrees a clean exit rather than inventing a smaller role, and opens a Topgrading search for the leader. The VP leaves on agreed terms, with an interim owner named and customers told the same week.
Dishonesty is not an outgrowing question: falsifying numbers or misleading customers or the board leads to a consequence the same week, including termination. And an executive who is missing because the company changed strategy under them deserves a reset of the seat's outcomes before any judgment about the person.
From the Golden Section mistakes list, each paired with the play that prevents it.
Root-causing the misses first separates a person problem from a goal, resource or incentive problem.
Delaying a decision about a leadership seat lets the gap spread into every team that depends on it.
Hoping an executive takes the hint prolongs the damage; the exit has to be clear, definitive and handled with dignity.
In the order we would run them. Each is on its own page, most with a free Excel template.
Defines the seat and its outcomes, so the judgment is about the role rather than the relationship.
Gives effort gaps dated steps, a written record and a decision at the end.
Clarifies scope, compensation and off-boarding terms before an exit conversation.
Uses the board as a sounding board and air cover for an unpopular but necessary decision.
Runs the replacement search against a scorecard written for the next stage.
Executive plays The plays for defining seats, handling underperformance and hiring leaders, which decide how cleanly a leadership change goes.
Decide on the evidence, review the agreement and equity terms with counsel, and name the interim owner before you speak to them. Have the conversation in person, early in the week, clearly and without hinting, and agree the terms and the message to the team. Then tell the leadership team, the board and affected customers the same week, so the story is yours rather than the rumor's.
Usually not. When the gap is scope, our default is a clean exit. Move someone only if the role is real, meaning it has outcomes the company needs and would hire for anyway. A title created to avoid a hard conversation delays the decision and usually ends in the same place a year later, with less goodwill on both sides.
For an effort gap, about 30 days of written expectations, then a plan of up to 60 days with a decision at the end date. For a scope gap, time does not help, so have the conversation as soon as the root cause is clear.
Reviewed by Dougal Cameron, CEO & Co-Founder on 2026-10-22. Golden Section observations are labeled separately from external benchmarks and illustrative arithmetic.