3 Principles of Good Incentives

Part 3: Setting incentives that benefit everyone.

Michał Poczwardowski

Feb 05, 2026

Setting incentives is manipulation. I’ve used them to steer behaviour, but sometimes in an unexpected direction.

You can’t stop people from gaming incentives, but you can choose the game.

If you read Part 1 and Part 2, you already know why incentives matter and why humans are wired to game them.

In this final Part 3, we’ll go through principles of choosing the right incentives.

Examples

At Basecamp, employees who have worked there for more than two years receive a share of the profits. They get one share per month, and every year the company shares 10% of their profit with employees. It rewards the company’s success and incentivises people to stay for the long term.

On Substack, writers want to monetise their newsletters, and Substack cuts a fee from paid transactions. The more writers earn, the more Substack earns. It’s win-win for both parties.

In my previous company, we had a set of metrics shared between all managers and across all business units. Revenue, gross margin, turnover and customer satisfaction all together. Dimensions of metrics balance it in a sustainable way. One team couldn’t just focus on one, which encouraged looking for alignment more often.

The Most Important Question

As we discussed in Part 1:

When you don’t understand people’s actions, look at their incentives.

Invert it, and we have:

Set incentives, and people’s actions will follow.

Which means: people consistently respond to incentives, even when they claim otherwise. This is one of the most important things to take away from the article.

Incentives shouldn’t exist without a clearly defined outcome of actions that are encouraged to achieve it. Which boils down to the most important questions:

What would you like to achieve?

The honest answer to this question is a baseline.

From that baseline, you can extract actions that support the goal. Only then, think what can motivate people to act.

Based on my experience, I’ve distilled 3 principles for setting incentives.

3 Principles of Good Incentives

Principles work for any kind of incentives: bonuses, salary brackets, promotions, recognition, or visibility.

1. Know Desired Outcome

Principle: Know what your desired outcome is, only then encourage actions that lead to the desired outcome.

Use second-order thinking: once you achieve this metric, then what? What are the consequences of focusing too heavily on it? Does it bring us closer to the desired outcome? What about things that may suffer because of it?

Remember the rats in Hanoi from Part 2? Paying for cut tails was just a proxy. It didn’t reduce the rat population, which was a desired outcome. Even requiring full dead rats wouldn’t have worked because people would just breed them. They focused on the wrong thing.

2. Stabilise with Multiple Metrics

Principle: Single metrics create tunnel vision, multiple ones create healthy tension between them.

It’s harder to game multiple balanced metrics.

For example: revenue alone won’t guarantee quality deals, but revenue + retention + satisfaction together will

3. Share and Align

Principle: Align incentives across groups, as shared principles mean shared direction.

It’s easier to operate when people understand what is the underlying goal, no matter which set of incentives will be used. As with example from Basecamp, where all incentives are available to the public and clear. Similarly, when people openly share what motivates them, it’s easier to find opportunities within the team.

Broadcast:

What you’re measuring
How it’s weighted
Why it matters

Testing via Questions

Before implementing any incentive system, answer these:

What behaviour are you trying to encourage?
What metrics support desired outcome?
How will people game the metrics?
What are second-order consequences?
What will this incentive cause people to stop doing?
Are these incentives aligned across groups?

If you can’t confidently answer all six, you still need iteration on your system.

Final Thoughts

The best incentive system is invisible, and people are driven by the work itself. But the reality is that I worked to a large extent for money, as the majority of the population probably does.

Design incentives as amplifiers, not something that is the only motivation to work.

Before setting them, understand what you want to achieve, stabilise them with multiple metrics and broadcast transparently.

You can’t stop people gaming incentives. But you can choose games worth playing.

The full series:

Part 1: Why Incentives Matter: When you don’t understand people’s actions, look at their incentives.
Part 2: The Cobra Effect: People will find the path of least resistance to win the game, and they are smarter than our metrics.
Part 3: Setting Incentives: Design incentives that align motivation with outcomes.

Thanks for reading,
— Michał

Share Perspectiveship