How to Design a Recruiter Incentive Plan That Changes Behaviour

A commission plan is the only strategy document in a staffing firm that everybody reads carefully. Whatever the leadership deck says about quality and partnership, recruiters optimise what they are paid for. Designing a plan properly means starting from the behaviour you want, pricing it, and accepting that the answer may cost more than the plan you can most comfortably afford.

Last reviewed August 2026. Practical guidance drawn from Sortinghat customer observation.

Start by writing down the behaviour

Before any numbers, list what you actually want a recruiter to do. Most firms have never done this and the list is short.

Fill roles profitably. Decline roles that will not fill. Build accounts that pay next year. Pass a role to a colleague better placed to fill it. Maintain the records so the desk survives their departure.

Then check each against the current plan and ask whether it is paid for. Usually the first is paid, the second is punished, and the remaining three are unpaid, which explains precisely what the team does.

What to pay on

Margin, where you can measure it. A flat percentage of billings makes a recruiter indifferent between a high-margin difficult placement and two easy low-margin ones. Margin-based plans fix the incentive and require data most firms have not built.

If you cannot measure margin by desk and by placement, fix that first. Redesigning a plan on unreliable data produces disputes rather than motivation, and it is why front and back office on one data model matters commercially, per where back office margin leaks.

Avoid paying on activity. Every activity count can be raised by lowering the standard, which is the argument in the recruiter productivity myth.

Over what period

The calendar month is the default and it causes most of the observable dysfunction.

A placement landing on the 2nd rather than the 28th is worth materially different amounts depending on where the recruiter sits against target, so start dates get influenced by commission rather than by the client. Near a threshold, deals are pulled forward or pushed back for reasons that have nothing to do with delivery.

A twelve-week rolling window removes most of this. It costs some simplicity and it stops the month-end behaviour that everybody pretends not to notice.

The four failures every plan has to handle

Ownership disputes. Who is credited when a candidate sourced two years ago by someone who has left is placed today. This is a compensation question, not a permissions question, and it needs answering before the software is configured, per candidate ownership rules.

Split placements. One person owns the client, another the candidate. Decide the split in advance or every case becomes a negotiation between two people who both want to win it.

Departures. What is paid on placements in flight when somebody resigns. Settle it in the employment contract rather than during an exit conversation.

Clawbacks. What happens when a placement falls through inside the guarantee period. A plan without a stated position produces an argument at the worst moment.

Paying for work that pays later

The hardest problem in commission design. Account development, referral cultivation and record maintenance all produce revenue later and cost billing time now.

Three approaches that work. A small explicit component separate from placement commission, which is the simplest. A retrospective account bonus paid when an account crosses a revenue threshold, which rewards the person who built it. Or protected time, which is not compensation but removes the cost.

What does not work is expecting it as professionalism. If it costs a recruiter money and earns them nothing, it does not happen, and the firm concludes people are short-termist rather than reading the plan correctly.

A design sequence

1. List the behaviours you want. Five or six, written down.

2. Check which the current plan pays for. Honestly, including the ones it punishes.

3. Confirm you can measure margin. If not, that is the prerequisite project.

4. Choose the period. Rolling beats calendar for most firms.

5. Write the four edge cases. Ownership, splits, departures, clawbacks.

6. Model the cost at three performance levels. Poor, expected and exceptional. A plan that is only affordable if nobody performs well is not a plan.

7. Show it to two recruiters before launching. They will find the gaming route in ten minutes, which is cheaper than finding it in a quarter.

Frequently asked questions

The exercise that takes an afternoon

List the six behaviours you want from a recruiter. Mark which ones your current plan pays for, which it ignores, and which it actively punishes.

Most firms find that declining a bad role and building an account that pays next year both fall in the third column. That is not a people problem, it is a plan problem, and it is fixable.