Alt text: A close-up of hands exchanging money for a basket of cobras across a desk, illustrating the parable of an incentive scheme that backfires

Second Look

Should Sales Bonuses Reward Revenue or Profit Per Deal?

Advancement Quest Team

July 28, 2026

There's a well-known story in economics, more famous as a parable than confirmed as fact, about colonial Delhi and a cobra problem. The government offered a bounty for every dead cobra handed in, and for a while it worked exactly as intended: cobras got killed, bounties got paid, the snake population fell. Then people worked out that breeding cobras was considerably easier than hunting them, and a small industry of cobra farming sprang up, each snake worth exactly as much dead as it would have been in the wild.

That's usually where the story gets told and stopped. But there is more. Once the administration noticed what was happening, it scrapped the bounty. With no more payout for a dead snake, the farmers had no reason to keep feeding a shed full of them, so they released them. The city ended up with more cobras than before the scheme ever existed.

Both halves of that story matter, and it's the second one that's usually left out - although it is as illustrative as the first one. A reward gets gamed, so the reward gets removed, and removing it creates a new problem nobody designed for - which is close to what happens when a business swaps a revenue bonus for a profit bonus or vice versa, expecting the new metric to simply be safer.

⚡ The metric you reward becomes the rule your sales team uses when nobody is watching.

A sales bonus is a set of instructions, whatever the paperwork calls it. Tell the team revenue is what gets rewarded, and revenue is what they'll optimise for, in whatever way is fastest and easiest to reach, not necessarily in the way that was intended when the target was set.

Take a deal priced at £100, costing £80 to deliver, for a profit of £20. Offer a 10% discount to get it over the line, and revenue drops to £90 while profit drops to £10. Revenue fell by a tenth. Profit fell by half. A salesperson rewarded on revenue still gets paid - less than a full-price deal, but more than the nothing they'd have earned if the sale hadn't closed. The business has just lost half its margin on that deal.

That's the visible version of the problem, and it's real. Chase revenue as the number that matters, and discounting becomes a rational, entirely predictable response, along with taking on low-margin work that helps the top line without helping the business, or chasing large contracts that look impressive on a dashboard and turn out to be expensive or difficult to actually deliver.

Switch to profit per deal instead, and the gaming doesn't stop. It just changes. A salesperson rewarded on margin has every reason to protect it by cutting corners on delivery or quality where the shortfall won't show up until later. Strategically useful but lower-margin work - a new-logo account, a foothold client worth having for reasons the bonus structure doesn't see - starts getting quietly avoided, because it drags the number down for no reward. And a deal's profit, measured today, has no way of accounting for whether the client stays five years or five months, so a bonus built entirely around this quarter's margin can end up working against the relationship's actual lifetime value.

⚡ Before choosing the target, check which behaviours it will reward - including the ones you do not intend.

Neither metric is the safe one. Poor-fit customers, unrealistic promises, and weak delivery can all happen under a revenue bonus or a profit bonus alike - the incentive doesn't create the underlying risk, it just decides which shortcuts feel worth taking. Whichever number ends up on the compensation plan, the actual design question is not "does this reward the right outcome," but "if someone optimised for exactly this number and nothing else, what would they do, and would we be comfortable watching them do it."

That question rarely gets asked before a metric goes live, and it's usually cheaper to ask it in a planning meeting than to find the answer six months later, in the numbers.