50%
wasted, on a good week
If you're spending money on driver incentives, there's a good chance most of that budget is doing nothing. In the worst cases, it's doing worse than nothing.
That's not a rounding error. In markets we've run, wasting half the incentive budget was a good week. Eighty percent wasted was typical. And yes — it is entirely possible to lose more than 100% of what you spend, where the incentive doesn't just fail to move behavior, it actively removes supply from the road when you need it most.
Here's why, and what actually works instead.
Incentives exist to change what a driver does — get them online during a dead hour, get them to take one more trip, get them to show up on the day your demand peaks. If a driver would have done the thing anyway, the incentive didn't do anything except hand over cash. That's the entire test. Most programs fail it without anyone noticing, because incrementality — what would have happened without the incentive — is genuinely hard to measure, and it's much easier to just look at total trips and assume the incentive worked.
The simplest incentive design is also the most wasteful: give every driver an extra €2 for every trip they complete. It feels fair, it's easy to communicate, and it's almost impossible to justify.
The problem is obvious once you say it out loud: a large share of the drivers taking that bonus were going to take the trip anyway. You have no way to tell, at the moment you pay it, whether you just moved someone's behavior or just handed a regular driver a tip for doing what they were already going to do. Because every driver gets treated identically regardless of how different they actually are — full-time versus part-time, morning versus night, high-earner versus low-earner — you're guaranteed to be overpaying a large fraction of your driver base for zero incremental supply.
The more sophisticated version looks smarter on paper: "Do 50 trips this week, get €100." It feels targeted. It isn't. It fails in three distinct ways, and all three are common enough that most operators running this kind of program are hitting all of them simultaneously.
The high-frequency driver takes it for free. A driver who normally does 60 or 70 trips a week blows past 50 without changing anything about how he works. He collects the bonus as pure profit. You paid for supply you already had.
The low-frequency driver never reaches it. A part-time driver doing 30 trips a week usually isn't going to jump to 50 just because there's a bonus attached — the gap is too large, or the reason he's part-time (another job, family, whatever) doesn't move because of a bonus threshold. So the incentive doesn't reach the driver it was theoretically designed to influence, either.
And the genuinely dangerous failure: it can reduce total driver hours. Picture a driver who normally does 60 trips a week because that's roughly what he needs to earn. If the threshold is 50 trips for a fixed bonus, he now hits his income target faster — he gets the bonus, and the marginal reason to keep driving past that point is gone. He stops at 50. You just paid him to work less.
The pattern across all of this is consistent. The more granular an incentive — targeted by driver, by hour, by the specific gap between what's happening and what you need to happen — the more of your budget actually changes behavior. The more blanket an incentive — same amount, same rule, applied across the board — the more of it you're simply giving away to drivers who were never going to change anything.
We're not going to lay out the exact targeting mechanics here — that's the part every operator has to build for their own market, and it's genuinely where the competitive edge lives. But the diagnostic question is one anyone can ask of their own program today: for any incentive you're currently running, can you say with confidence which drivers it actually moved, versus which drivers would have done exactly the same thing for free? If the honest answer is "we don't really know," you're very likely funding the 80% that does nothing — or worse, the sliver that's actively working against you.
Driver incentive design is arguably the single highest-leverage lever in ride-hailing economics, and it's one of the least understood — because it's technical, because the failure modes are invisible in aggregate trip numbers, and because "we're spending money on drivers" feels like it should be working even when it isn't.
If you're a founder-led operator competing against a much larger, better-funded incumbent, this is not a nice-to-have optimization. It's one of the few places where a smaller, faster-moving team can outperform a much bigger budget — not by spending more, but by spending what you have on the drivers who actually respond, instead of the ones who were driving anyway.
A short working call — I'll tell you where your spend is actually leaking.
Book a working call