Is a long pickup Uber ride worth it? The deadhead is what kills good-looking offers

The offer card puts the pickup distance in small type at the top, above the payout, and most drivers read it the way you read a weather forecast. It is the number that decides more of your week than the payout does.

I let a $19.80 offer time out last week and took an $11.40 one two blocks later. The difference was entirely the pickup, and once I show you what a pickup actually costs, that call stops looking eccentric.

The short answer

Pickup miles cost you at your full cost per mile and pay you nothing. Pickup minutes stretch your hourly rate without adding a cent to the fare. It hits the numerator and the denominator at the same time, which is why a long pickup can gut an offer without the payout figure changing at all.

I pass on anything over 6.0 miles or 12 minutes of pickup, unless the offer still clears 1.5 times my floor after every cost is counted. That is the whole rule. The rest of this article is why those numbers, and where they are wrong.

Two identical offers that are not identical

Here is the cleanest way to see it. Same payout, same trip, different pickup.

An offer of $16.20, trip 8.4 miles and 22 minutes. My cost per mile is $0.26 and my floor is $20 an hour net.

With a 1.6 mile, 5 minute pickup: Total miles 10.0. Cost $2.60. Net pay $13.60. Total time 27 minutes, or 0.45 hours. Net hourly: $30.22.

With a 6.8 mile, 16 minute pickup: Total miles 15.2. Cost $3.95. Net pay $12.25. Total time 38 minutes, or 0.633 hours. Net hourly: $19.34.

Same money on the card. Same rider going the same distance. One is half again over my floor and one is under it. Nothing about the payout tells you which one you are looking at.

That spread, more than ten dollars an hour, is what the pickup does. And it does it quietly, because the payout is the number the card puts in the biggest type.

What the deadhead actually costs

Break down that 6.8 mile, 16 minute pickup.

The miles: 6.8 at $0.26 is $1.77 of fuel and wear, spent before the fare starts. On the fuel half alone, at EIA's published national average of $4.096 a gallon for the week of July 27, 2026, a 28 MPG car burns about $0.146 a mile, so roughly $0.99 of that is gasoline you are buying to arrive somewhere empty.

The minutes: 16 minutes of your working time. Against a $20 an hour floor, that is $5.33 of value.

Together, about $7 of real cost for the privilege of starting the ride. None of it appears anywhere on the offer card, and no summary at the end of the week will separate it out for you either.

Now stack that across a shift. Six long pickups in an evening is roughly an hour and a half of unpaid driving and forty-odd miles of wear. That is not a rounding error. That is most of a paid trip's worth of time and cost, spent arriving.

Levi Spires and the 8 minute finding

The most useful public data I know of on this comes from a driver rather than a company.

Levi Spires documented an experiment in which he declined 184 Uber rides and tracked what happened. Two things in it are worth carrying. First, his hourly rate swung roughly threefold across the period, from around $18 to around $54 an hour, with no surge indicator shown in the app to explain the difference. Second, he identified a pickup over 8 minutes as the strongest single predictor of a bad trip in his data.

That second finding is the reason I take the pickup seriously as a filter rather than as one input among many. It is not the payout, not the trip distance, not the time of day. It is the pickup.

It is also one driver, one market, one stretch of driving. n=1. I am not going to restate it as what every driver should expect, and neither should anyone else. What it is worth is a hypothesis you can test in a week of your own data, which is a much better use of it than treating it as a rule.

Why my limit is 12 minutes and not 8

Reasonable question, given the above.

Spires found 8 minutes to be the sharpest dividing line in his own data. My veto sits at 12 minutes and 6.0 miles. That is deliberate, and there are two reasons.

The first is that the veto does not work alone. It sits on top of the full arithmetic, which has already subtracted the pickup miles and already added the pickup minutes to the denominator. An offer with a 10 minute pickup that fails on economics gets rejected by the maths before the veto ever looks at it. The veto is there to catch the ones that pass the arithmetic but carry risk the arithmetic cannot see, which is a narrower job.

The second is market density. Durham is not a dense city. An 8 minute limit here would rule out a large share of everything that comes in, and a filter that rejects most of your volume is not a filter, it is a decision to stop working. In Manhattan the same limit would be almost costless. The right number is a function of your own market, and if I published 8 as a universal rule I would be handing drivers in spread-out markets a rule that quietly loses them money.

Set your own. Just have one, and apply it the same way every time.

Why a hard limit at all, when you already have the arithmetic

This is the part I had to be argued into, so I will make the case properly.

If the hourly calculation already accounts for the pickup, a separate distance and time veto looks redundant. It is not, because the calculation assumes the trip goes as described, and long pickups are exactly where that assumption breaks.

The rider cancels while you are eight minutes out. Now you have driven those miles and spent those minutes for a cancellation fee that does not cover either.

Traffic on the way to a pickup is unpaid in a way traffic on a trip is not. A trip that takes longer than estimated at least pays some time component. A pickup that takes longer than estimated pays nothing at all.

The drop-off compounds. A long pickup often means you were pulled out of the area you were working, and the trip then leaves you somewhere else again. You paid the deadhead twice, once to get there and once in the dead time after.

None of that is in the offer card's numbers, and none of it is in the hourly calculation. That is what the veto is for.

The one exception

A genuinely excellent offer should not be thrown away by a rule of thumb, so the veto has an override: if the offer still clears 1.5 times my floor after every cost and all the pickup time, it stands.

At a $20 floor that means $30 an hour net. Worth seeing what actually qualifies.

Clears it. Offer $31.80, trip 15.2 miles and 27 minutes, pickup 7.4 miles and 15 minutes. Total miles 22.6, cost $5.88, net $25.92, total time 42 minutes or 0.7 hours. Net hourly $37.03, which is 1.85 times the floor. The pickup trips both limits and the offer takes it anyway.

Does not clear it. Offer $23.40, trip 16.8 miles and 29 minutes, pickup 7.1 miles and 15 minutes. Total miles 23.9, cost $6.21, net $17.19, total time 44 minutes or 0.733 hours. Net hourly $23.44, which is 1.17 times the floor. That is a perfectly decent offer that clears the floor comfortably, and it still does not earn the exemption. It goes.

The gap between those two is the whole point of the override. It is not there to let good offers through. It is there to let exceptional ones through, and a few percent over the floor is not exceptional.

Where the pickup number comes from, and why to distrust it

One more practical note. The pickup distance on the card is a straight-line or routed estimate, and the pickup time is an estimate on top of an estimate. Neither accounts for the apartment complex with one entrance on the wrong side, the hotel where the rider is on the fourth floor, or the airport queue.

My working assumption is that a displayed pickup runs longer than shown more often than it runs shorter. That asymmetry is another reason the veto sits where it does rather than tighter to the arithmetic. If the estimate were unbiased I would trust the calculation alone. It is not, so I do not.

Doing this in the seconds you actually have

Nobody is dividing 12.25 by 0.633 with a countdown running. What you can do is train the shape of it.

Read the pickup first, before the payout. If it is over your limit, you are done, and you never had to do arithmetic at all. If it is under, then read the payout and run the rough version: all the miles at about a quarter each, all the minutes including the pickup, and see whether the result is anywhere near your floor.

That ordering is the single habit change I would recommend to any driver reading this. The card is designed to be read payout first. Read it pickup first and a large share of the traps disappear before you have engaged with them.

The full method is in the 20-second math I run on every offer, the cost per mile that everything above depends on is in what a mile actually costs you, and the fuel half of that number varies enough by vehicle to move the line, which is worked out in the best car for Uber drivers.

Passing on offers is not free either, and the honest accounting of what it costs is in what declining actually does to your account. You can run any single card through the calculator without installing anything.

DriverSignal adds the pickup you were about to forget, subtracts your own cost per mile, and shows the net and the net hourly before the card times out. It never touches Accept or Decline. But the habit matters more than the tool. Read the pickup first, and know your worth before the ping lands rather than after.