Never let your client tie your pay to factors you cannot control
In service negotiations, clients frequently push for performance pricing, wanting to pay only if their overall revenue increases. Nobel laureate Bengt Holmström showed why this breaks deals through the informativeness principle: contracts should strictly tie compensation to performance measures the provider directly controls. Linking your payout to a client's bottom line penalizes you for their poor marketing or broader market downturns. Insist on tying incentives exclusively to operational metrics your solution manages.
The Core Dilemma of Pay and Performance
In commercial agreements, parties routinely encounter a fundamental friction known in economics as the principal-agent problem. One party, the principal, hires another party, the agent, to perform a task. If the agent's effort were immediately visible and easily documented, the contract would be straightforward: the principal would simply pay for verified effort. In practice, the principal cannot perfectly observe the agent's actions, leading to moral hazard. Because the buyer cannot see how much diligence or care the provider actually exerts, the buyer naturally attempts to base payment on final outcomes, such as end revenues or profits.
While shifting financial risk to the provider seems like a tidy way to ensure motivation, it introduces serious structural flaws. Final business results rarely reflect individual effort in isolation. A project's end outcome is shaped by an intricate combination of the provider's work, the client's internal operations, and random external shocks. By demanding that pay depend entirely on bottom-line figures, a client forces the provider to absorb risks that have nothing to do with how well the hired work was executed.
The Informativeness Principle
In the late 1970s, economist Bengt Holmström developed a foundational mathematical framework to explain how contracts should handle uncertainty and incentives. His central finding became known as the informativeness principle. Holmström demonstrated that an optimal contract should incorporate every performance measure that provides information about the actions the agent took, but must exclude any measure that fails to clarify those actions.
Under this principle, linking compensation to a metric is only sensible if that metric reveals something specific about the provider's diligence. If an outcome indicator is heavily influenced by factors outside the provider's direct scope—such as general macroeconomic swings, the client's supply chain delays, or their internal sales execution—it adds statistical noise rather than relevant data. The informativeness principle proves that tying pay to noisy, uncontrolled metrics weakens the contract, because it penalizes or rewards the provider for events they could neither cause nor prevent.
The Trade-off Between Incentives and Insurance
A central mechanism in Holmström's analysis is the inescapable trade-off between providing incentives and offering insurance against risk. In typical contractual relationships, the party doing the work is treated as risk-averse, meaning they value stability and suffer disproportionately when exposed to arbitrary financial volatility. The hiring organization, by contrast, is usually larger and better positioned to diversify across multiple projects and market fluctuations.
When an agreement ties remuneration strictly to bottom-line results, it exposes the risk-averse provider to severe outcome uncertainty. To accept that exposure, a rational provider must demand a substantial risk premium—a higher overall expected payout to compensate for the chance of unearned losses. Conversely, if the client offers a purely fixed fee with zero link to results, the provider bears no risk, but the client retains no contractual leverage over hidden effort. Holmström demonstrated that an effective contract carefully balances these forces, introducing incentive pay only where the metric cleanly tracks the agent's actions without drowning them in external risk.
Filtering External Noise Through Comparative Metrics
Holmström extended this logic to explain how contracts can systematically strip out broader market randomness through relative performance evaluation. If an entire industry experiences a downturn due to raw material shortages or sudden regulatory shifts, judging an individual provider solely against raw financial yields unfairly punishes them for broad environmental distress. Conversely, an undeserving provider might receive windfall gains during an effortless market boom.
To resolve this, Holmström showed that contracts become more efficient when performance is benchmarked against peers or broader sector indexes. If a firm's competitors suffer an average decline of twenty percent, an agent who limits their client's decline to five percent has clearly delivered superior effort, even though the absolute performance was negative. By filtering out common shocks shared across an industry, contracts isolate the signal of genuine performance from systemic noise, ensuring that compensation reflects actual competence rather than random luck.
The Distortion of Multi-Tasking
In later research conducted with Paul Milgrom, Holmström analyzed what happens when a role requires balancing multiple distinct activities. In almost every professional engagement, a provider must divide their energy between tasks that are easy to quantify and tasks that are difficult to measure, such as maintaining long-term quality, documentation, or client goodwill. When strong incentives are tied exclusively to the measurable metrics, rational agents redirect their focus entirely toward the rewarded tasks while neglecting the unmeasured ones.
This multi-tasking model highlights why broad outcome-based bonuses frequently backfire. If an outside vendor is evaluated solely on immediate transaction volume, they have an economic incentive to sacrifice long-term relationship stability or system architecture to hit short-term targets. In scenarios where critical dimensions of work cannot be measured precisely, Holmström and Milgrom demonstrated that optimal contracts must intentionally weaken incentive pay, relying instead on fixed fees to prevent perverse behavior.
Incomplete Contracts and the Role of Control Rights
Contract theory was further transformed in the mid-1980s by Oliver Hart, who shared the 2016 Nobel Prize with Holmström. While Holmström analyzed how to design pay around observable signals under moral hazard, Hart examined the reality of incomplete contracts. In complex, long-term business dealings, it is impossible to write down every future contingency, unforeseen event, or operational dispute in advance. Because contracts cannot foresee every eventuality, writing comprehensive outcome formulas is inherently limited.
Hart’s work showed that when contracts are inevitably incomplete, the critical question is not merely how money is split, but who holds residual rights of control—the legal authority to make decisions when unforeseen circumstances arise. If a client retains full control over product pricing, marketing budgets, and operational staffing, the provider has no authority to correct poor execution. Forcing a provider to stake their earnings on an outcome while denying them the decision rights necessary to shape that outcome contradicts the core economic logic of contract design.
Key takeaways
•The informativeness principle establishes that compensation should only reflect metrics that provide clear information about the provider's actual actions, excluding irrelevant market noise.
•Tying pay to metrics beyond an agent's control forces them to absorb unmanageable risk, which rationally requires a higher risk premium or creates contractual breakdown.
•Rewarding only easily quantified outcomes leads to multi-tasking distortion, incentivizing agents to hit narrow measured targets at the expense of unmeasured long-term quality.
•Under incomplete contract theory, responsibility for final outcomes cannot be separated from decision rights; providers should not accept financial liability for variables they lack the authority to govern.