If you've replaced the same role three times, the problem isn't talent. It's your design. Most underperformance is not a people problem. It's a structural problem that keeps consuming people until someone decides to fix the structure instead of the occupant.
The instinct to replace is understandable. Someone is not performing. The work is suffering. The team is frustrated. The path of least resistance is to conclude that the person in the role is the wrong person, hire someone new, and move on. This feels decisive. In many cases it is just expensive.
I have seen organizations replace the same role four and five times in two years, each time with people who were demonstrably capable, each time producing the same outcome. The third time this happens, the talent hypothesis has already failed. The problem is structural. But the structural explanation is harder to accept than the personal one, because it requires the organization — and specifically the leaders who designed the structure — to own the failure rather than transferring it to whoever currently holds the title.
Roles that are too broad. The most common structural failure is a role that exists to catch everything that has not been assigned anywhere else. These roles accumulate scope over time, rarely by design. When someone new joins, they discover that the job description and the actual job are completely different things — that the role spans four or five distinct function areas, none of them resourced adequately, all of them expecting full ownership. The person cannot succeed. The role is not designed for success. It is designed to absorb overflow.
Roles that are too reactive. Some roles exist to respond to demands rather than to produce defined outcomes. The person in the role is never sure whether they are succeeding because the role has no clear definition of success — only an expectation that requests will be handled quickly. This produces exhaustion masquerading as underperformance. The person is often working harder than almost anyone else on the team and showing nothing for it, because the work itself has no accumulating output. Every day starts at zero.
Roles measured by activity, not outcomes. When performance is evaluated on the basis of visible effort — meetings attended, emails processed, tasks completed — the role incentivizes busyness over impact. High performers find this profoundly unsatisfying. They want to know that their effort moved something important. A role that cannot tell them whether their effort mattered will not retain them for long.
Roles owned by many, accountable to none. Shared ownership is a common solution to cross-functional work, and it almost always produces the same outcome: nobody owns it. When a decision requires sign-off from multiple stakeholders and each of them has partial authority, the person nominally responsible for the outcome has no actual power to produce it. Accountability without authority is not accountability. It is a setup for failure with organizational deniability built in.
People don't fail in well-architected systems. They fail in confusing ones. The best leaders diagnose before they replace — and fix structure before they blame people.
Before you open a requisition, ask four questions about the role: What is the one thing this person is accountable for producing? What authority do they have to produce it? How will they know whether they are succeeding? And who, besides them, shares ownership of the outcome?
If you cannot answer the first question in one clear sentence, the role is not designed. If you cannot answer the third question with a measurable benchmark rather than a vague impression, the performance management system is not designed. If the answer to the fourth question is "several people," you have not designed a role — you have divided a problem and hoped someone will absorb the cost of the seams.
The leaders who build organizations that develop and retain excellent people are not necessarily better at identifying talent. They are better at designing the conditions in which talent can succeed. Talent is common. Well-architected roles that enable talent to produce and grow are significantly less so. The constraint is rarely the people. It is almost always the system the people are being asked to operate in.
The pattern that consistently produces the worst outcomes is adding headcount before adding clarity. When an organization is struggling with performance and the first response is to hire more people, the new people inherit the same structural problems that defeated the previous cohort — but now there are more of them, compounding the confusion and increasing the cost of the eventual diagnosis.
Clarity costs almost nothing. A clear definition of what a role produces, what authority it carries, and how performance will be measured can be written in an afternoon. It changes what a hiring conversation sounds like, what onboarding covers, what the first ninety days are built around, and what performance review conversations focus on. Everything downstream of that definition gets easier. Everything downstream of its absence gets harder — until eventually it gets expensive.
Design before you replace is a diagnostic principle: before replacing an underperforming person, ask whether the role itself is well-designed. Sterling Grey argues that most repeated hiring failures — where the same role turns over two or three times — are evidence of a structural problem: unclear accountability, impossible scope, conflicting reporting lines, or misalignment between what the role is supposed to do and what the organization actually needs. Replacing the person without fixing the structure just resets the clock.
The four most common role design failures are: unclear ownership (two people share responsibility for the same outcome, so neither fully owns it), impossible scope (the role requires capabilities or bandwidth that no single person could provide), structural conflict (the role reports to one part of the organization but is accountable to another), and mismatched expectations (the organization needs one thing but the role description promises another). Each of these will produce consistent underperformance regardless of who holds the role.
The diagnostic starts with two questions: Have multiple people struggled in this role in similar ways? And can you describe, in specific and measurable terms, what success in this role actually looks like? If the answer to the first question is yes and the second is unclear, the role has a design problem. The next step is to map the role's accountability, decision rights, and success criteria explicitly — often revealing the structural issue that has been consuming people.
The Velocity Gap Assessment identifies the structural patterns — in roles, accountability, and decision rights — that are creating performance friction and consuming talent your organization cannot afford to lose.