When I started working with Carbon Talent, the recruiting firm founded by Lali Devaki, the team was performing on hustle and founder energy. Everyone worked hard. The output was inconsistent. The same kind of search would take dramatically different amounts of time depending on who was running it, and the variance was not explained by search difficulty.
The intervention was not about the people. It was about the system. We built clarity around ideal client profile, defined roles more explicitly, and created a lightweight process that made the implicit steps in a successful search explicit and repeatable. Within two quarters, output variance dropped significantly and the team could take on more work without proportional headcount growth.
If you replaced the team with equally capable people tomorrow, would the output be different six months from now? If the honest answer is no, the problem is the system.
High Performance Is a System Property
The instinct when a team underperforms is to look at the people. Sometimes that is the right diagnosis. More often, the people are capable and the system is working against them. Unclear priorities, conflicting incentives, missing information, process overhead that adds no value - these produce the same symptoms as capability problems and require entirely different interventions. (see also: misaligned incentives are almost always behind misaligned teams)
The diagnostic question: if you replaced the team with equally capable people tomorrow, would the output be different six months from now? If the honest answer is no, the problem is the system. Changing the people does not fix a system problem. It restarts the clock on the same problem with new faces.
The system properties that most consistently explain underperformance: unclear ownership of decisions, misaligned incentives between what the team is measured on and what the product needs, insufficient information flow between the team and the customers they serve, and coordination overhead that consumes the time needed for actual work.
Clarity Is the Underrated Performance Driver
The factor that most consistently separates high-performing teams from average ones is not skill level, culture, or tools. It is clarity. Specifically: clarity about what success looks like, clarity about who owns which decisions, and clarity about what the team is explicitly not trying to do.
The last one is hardest. High-performing teams have made explicit choices about what they are not building, not optimizing for, not prioritizing. These choices are as important as the positive priorities and they are almost always absent from the teams that struggle with focus. (see also: most decisions are two way doors)
The clarity test: ask each person on the team to write down, independently, the top three things the team is trying to accomplish this quarter and the top two things the team has explicitly decided not to prioritize. Compare the answers. The degree of variance tells you how much shared clarity actually exists.
Trust Is Built in Specific Moments
Team trust is not built through trust-building exercises or off-site retreats. It is built in specific moments where someone takes a risk and the outcome either validates or punishes it.
A team member flags a problem early and the response is gratitude rather than blame. A leader admits uncertainty rather than projecting false confidence. An engineer pushes back on a timeline and the pushback is taken seriously. These moments accumulate into an organizational memory that shapes how much risk people are willing to take.
You can rebuild trust after it has been damaged but it takes time and consistent behavior, not a single intervention. The organization remembers how previous risk-taking was received.
Trust is built in specific moments where someone takes a risk and the outcome validates or punishes it. These moments accumulate into an organizational memory that shapes how much risk people are willing to take.
What Accountability Actually Requires
Accountability without authority is frustration. The teams where accountability is genuinely high share one structural property: the person accountable for an outcome has meaningful control over the inputs that determine it. When accountability and control are misaligned, the result is either learned helplessness or perverse behavior. (see also: accountability is a team sport)
Building accountable teams requires first being honest about who actually controls what. Then aligning accountability to control, or changing the structure so the control follows the accountability. The organizations that resist this usually have a politics problem rather than a performance problem.
The accountability conversation that actually works: 'Here is what you are responsible for. Here is what you have control over. Here is what you need from others to do this well. Let's make sure the support you need is real before we hold you to the outcome.'
Measurement That Develops Rather Than Polices
The metrics a team tracks shape behavior more powerfully than any values statement. A team that tracks velocity will optimize for velocity. A team that tracks customer outcomes will optimize for customer outcomes. These are not the same thing.
The metric I push teams toward: rate of learning. How quickly is the team generating and validating hypotheses about what customers need? This captures velocity in the dimension that matters - not features shipped but understanding gained. It also resists the gaming that velocity metrics attract.
Measurement used for policing produces teams that protect themselves by managing the metrics rather than managing the product. Measurement used for development produces teams that improve.
