What Gets Rewarded: What Incentive Structures Reveal About What Organizations Actually Value

The Instinct

Every organization has a theory of what good performance looks like. It lives in job descriptions, performance review criteria, and compensation frameworks. It gets discussed in calibration sessions and communicated in annual reviews. It is documented, approved, and periodically revisited.

It is also, in most institutions, designed by people who are not the same people living inside it.

Compensation, incentive, and performance frameworks (collectively, "rewards frameworks") are built with genuine intent. Reward the right behaviors. Penalize the wrong ones. Align what individuals pursue with what the organization needs. The logic is sound. The execution is where things get complicated.

The complication is not that the frameworks are wrong. It is that they are designed without accounting for where else incentives, goals, and rewards are being defined. Risk appetite is set in one conversation. Compensation is designed in another. Strategic priorities are established somewhere else entirely. Each framework is internally coherent, but together, they may be pulling in different directions. And the people operating inside them will follow wherever the compensation points, regardless of what the other documents say.

The Gap

If you want to understand what an organization actually values, do not read its mission statement. Read its performance review criteria. Read its bonus structure. Read what gets celebrated in all-hands meetings and what gets quietly ignored in calibration sessions. Those documents are the revealed truth of organizational priorities, and they are frequently more honest than anyone intended them to be.

The manager who hits revenue targets while quietly accumulating risk is not acting in bad faith. They are responding rationally to what they were told mattered. The function that closes issues faster than anyone else without changing the conditions producing them is not being cynical. Speed of closure is on the scorecard. Root cause resolution is not. The team that adopts AI enthusiastically without anyone asking who owns the outcome is not being reckless. Adoption was the metric. Accountability was not.

In each case, the rewards framework did exactly what it was designed to do. It shaped behavior. The problem is that it shaped behavior toward the visible metric and away from the less visible but more consequential outcome sitting just behind it.

This is where the bystander dynamic becomes most concrete. Rewards frameworks are designed by consensus, informed by competing priorities, and approved by people who each bring a legitimate but partial view of what the framework should accomplish. Each party does its part. Nobody owns the question of whether the framework, as a whole, is pointing the organization in the right direction. That question falls into the space between functions, between conversations, between the people who set the targets and the people who absorb the consequences of hitting them.

The result is a system that rewards the production of visible success while the less visible costs accumulate elsewhere. The revenue target gets hit. The risk accumulates in the portfolio. The issue log clears. The underlying conditions persist. The AI adoption announcement goes out. The accountability question goes unanswered. Each outcome looks like success from one angle. From another, it is simply deferred exposure.

The dynamic is not limited to what happens inside institutions. Markets reward the announcement of cost reduction without asking what the cost of arriving at that point was. A company that lays off a significant portion of its workforce is celebrated for its discipline. The decisions that produced the overstaffing in the first place receive considerably less scrutiny. The scorecard resets at the moment of the announcement. Everything before it disappears.

This is the incentive at its most external and most visible. If the market rewards correction more than it penalizes the original error, the rational response is to make the error, then make the correction loudly. The accountability that should travel with the original decision does not. It arrives, if at all, only when the correction becomes unavoidable.

The Better Question

The first question a reward framework should answer is not what we want people to do. It is what do we want the organization to look like in three years, and are the behaviors we are rewarding today likely to produce that outcome.

That is a harder question to answer than it sounds. It requires the people designing reward frameworks to have a view on risk, on strategy, on operational resilience, and on the second and third order effects of the behaviors they are incentivizing. It requires the people setting risk appetite to have a view on what is actually being rewarded on the other side of the organization. And it requires someone to be asking whether those two things are aligned, on behalf of the institution rather than any individual function.

Most organizations have nobody in that role, and the matrix structure makes it genuinely difficult to assign. But the absence of a single owner cannot become an excuse for the absence of any owner. The reconciliation points where these frameworks intersect need named, specific accountability: people who are explicitly responsible for asking whether they are aligned, who have authority to act when they are not, and who absorb some consequence when they fail to close the gap. The bystander dynamic is not a failure of intent. It is a feature of the design that has to be deliberately countered.

A better framework measures outcomes, not just outputs. Not how many issues were closed, but whether the conditions producing them have changed. Not whether AI was adopted, but whether the organization is safer and more capable because of it. Not whether the revenue target was hit, but whether the growth was sustainable and the risk was priced correctly. And not whether the headcount was reduced, but whether the organization is actually more efficient, or simply smaller and hiring again in eighteen months. The difference between an output and an outcome is the difference between what the scorecard shows and what the organization actually achieved.

It also measures over time rather than in isolation. A function that consistently shows improvement against a hard baseline is telling you something different than one that hits a target in a good quarter. Proximity to an ideal, measured across multiple periods, is more informative than a single data point that may reflect timing as much as capability.

And it connects the metrics that are currently designed in parallel. Risk appetite and similar frameworks are often oriented around failure: what the organization wants to avoid, where the boundaries are, what constitutes an unacceptable outcome. Rewards frameworks are oriented around success: what the organization wants to achieve, what gets rewarded, what gets celebrated. Each orientation is legitimate. Together, they create a system where people are rewarded for outcomes on one side of the ledger without being held accountable for what those outcomes cost on the other.

If risk appetite and compensation are pointing in different directions, one of them is wrong. Finding out which one requires someone to look at both simultaneously, on behalf of the organization rather than either function. That is not a governance committee's job in the abstract. It is a specific responsibility that needs to be assigned, owned, and held to account.

The instinct is to design rewards frameworks that reward success. The better question is whether the definition of success embedded in those frameworks reflects what the organization actually needs, or simply what is easiest to measure.

Show me the scorecard. It will tell you everything.