Go to any gig driver subreddit or Facebook group on a weekday evening and you'll find the same post over and over: a screenshot of an offer, and three words — "Is this worth it?"
Scroll the replies and you'll see forty people arguing. "Easy accept." "Instant decline." "I'd take that all day." "That's insulting." Same screenshot. Same numbers. Completely opposite answers.
Here's the thing: most of them are right. They're just right about their own car, their own gas station, and their own definition of a good hour.
An offer screenshot doesn't contain enough information to answer the question. The payout and the mileage are on the screen. Your fuel economy, your local gas price, how much car you're spending per mile, and the hourly rate you're actually trying to hit are not — and those four things decide the answer.
The Same Order, Three Drivers, Three Different Answers
Let's take a real offer and run it for three drivers. The order is identical for all of them:
📱 The offer on the screen
$25.59 an hour. That number is the same for every driver on earth, because it's just payout divided by time. It's also the number that starts every argument, and it's the least useful number on the page.
Now here are the three drivers. Nothing about the order changes — only who's looking at it.
🌴 Maria — Sacramento, CA · 2019 Prius, still on payments
🎸 Dale — Knoxville, TN · 2016 Civic, paid off
🤠 Rob — Houston, TX · 2015 F-150, paid off
Three drivers. One order. One says take it, one says skip it, one says maybe. If all three were in the same comment thread, they'd each think the other two were wrong — and all three would be giving honest advice based on their own situation.
Look at how narrow the gap between them actually is: $21.34, $23.32 and $21.89 — about two dollars an hour between the thriftiest car and the thirstiest. The verdicts aren't two dollars apart. They're opposites. That's because the number that decides it isn't the car at all. It's the line each driver drew before the offer came in.
And notice who came out worst. Maria's Prius burns 18 cents of a gallon on this run — the least fuel of the three by a mile. She still gets the only flat "no," because she's paying off a newer car and she's set a Sacramento-sized target. The most efficient car in the group loses.
🧮 Stop Asking Strangers — Run Your Own Numbers
Put your MPG, your gas price and your target in once. The calculator saves them in your browser, so from then on you just type the payout, miles and minutes from the offer screen and get your answer in about five seconds.
Open the Order Calculator →The Part That Surprises People: Gas Isn't the Big Variable
Look at the gas numbers again. Maria is paying $4.85 a gallon in California. Dale is paying $2.85 in Tennessee — two dollars a gallon cheaper. And on this order, Maria's gas cost is seven cents more than Dale's.
Her Prius erased the entire California penalty. Meanwhile Rob, sitting on some of the cheapest gas of the three, pays nearly double what either of them does — because he's hauling a pickup truck.
On a single short order, your vehicle matters more than your state. A thrifty car in an expensive state beats a thirsty car in a cheap one. If you're comparing yourself to someone online, the car in their driveway tells you more than the state in their flair.
Where the state difference does show up
One order at a time, the gap is pocket change. A full week of driving is a different story. Here's the same three drivers at 100 miles a day, five days a week:
⛽ Fuel cost per week — 500 miles
Now compare apples to apples. Put Dale's exact Civic in Sacramento and his weekly fuel bill goes from $44.53 to $75.78 — the state alone costs him about $1,625 a year for identical work. So yes, geography matters. It just matters across a year, not across one screenshot. If you want to dig into this properly, our guide on fuel cost per mile walks through how to work out your own figure.
The Variable Nobody in the Comments Can Guess
Gas and MPG are at least knowable. Someone could ask you and you could answer. The third variable is the one that actually decided all three verdicts above, and it isn't a fact about your car at all — it's the hourly rate you've decided is worth leaving the house for.
That number is personal, and it should be. It depends on:
- What else you could be doing. If the alternative is sitting in a parking lot watching an empty map, a mediocre order beats zero. If your market is busy and a better offer lands every four minutes, your floor should be much higher.
- Your cost of living. $18/hr is a reasonable living in parts of Tennessee. In Sacramento it isn't. Maria's higher target isn't her being picky — it's her rent.
- Whether this is your income or your extra. Someone covering a car payment with gig work has a different floor than someone doing two hours after their day job.
- How much of your car you're willing to spend. Every mile is wear you'll pay for later, just not today.
Nobody replying to a screenshot knows any of that. They're answering with their number and calling it the number.
Why Gas Alone Isn't the Whole Cost
Gas is the only cost you actually feel on the night, which is why it's the only one most people subtract. But the mile also spent some of your tires, some of your brake pads, some of your next oil change, and some of what your car would have sold for. You pay all of that later, in one unpleasant lump, usually on a day you weren't planning to spend money.
That's the second number the calculator asks for, and it asks it as a question you can actually answer:
🔧 Your car's wear & tear
One question, because one thing explains most of the difference between drivers: how much value your car still has left to lose. A 200,000-mile Civic has already done its depreciating — what's left is mostly maintenance, and maintenance costs roughly the same for everybody. A car you're still paying off is shedding real money every mile, and that's the bill that actually hurts.
What's deliberately not in that number
Insurance, registration and your car payment aren't in those rates, and leaving them out is the whole point. You pay them identically whether you take this order or sit in the parking lot all night. A cost you can't avoid can't help you decide. Put it in the per-mile number and every long order starts looking unprofitable, when what's really happening is that you're charging one order for a bill that was already coming.
And about that 72.5¢ figure
You'll see drivers in those same threads subtract the IRS standard mileage rate — 72.5¢ a mile — and conclude that almost nothing is worth taking. Two reasons that goes wrong:
Roughly a third of it is costs you'd pay anyway. The rate is built to cover insurance, registration and finance charges alongside gas, depreciation and maintenance. As a tax deduction that's entirely correct. As a tool for deciding whether to tap "accept," it's answering a different question.
Gas is already inside it. The IRS rate includes fuel, so subtracting your gas and the full 72.5¢ charges you twice for the same gallon. Our rates above cover depreciation and maintenance only, which is exactly why the calculator can subtract gas separately using your MPG and your pump price.
Keep claiming the 72.5¢ at tax time — it's usually the better deduction, and for most gig drivers it's worth more than their actual expenses. Just don't use it to judge an offer. (We write about driving, not taxes. Talk to someone who does taxes for a living before you file.)
The Return Trip Nobody Screenshots
One more reason those comment threads go sideways: the app shows you one direction. It doesn't show the drive back.
On a normal single order, you end up roughly as far from your zone as you drove, so doubling the mileage is close enough. But on a stacked run it can be wildly different. A 14-mile triple might be seven miles out, four miles to the second drop, and two to the third — leaving you about three miles from where you started, not fourteen. Someone eyeballing that screenshot and doubling the number will tell you it's a terrible order. It might be a great one.
This is why the calculator lets you set the return distance yourself instead of assuming. It defaults to mirroring the outbound leg and you override it when the drive back isn't the same.
What to Do Instead of Posting the Screenshot
There's nothing wrong with asking — gig work is isolating and the groups are genuinely useful for things like which stores to avoid and how batching works. But for "is this worth it," you'll get a better answer in less time by doing it yourself:
1. Your MPG. Not the sticker. Fill up, reset the trip meter, drive a normal shift, divide miles by gallons at the next fill-up. Real city delivery mileage runs well below the highway rating.
2. Your gas price. Whatever you actually paid last time, not the national average.
3. Your target. The hourly rate that makes the shift worth doing. Be honest, and remember it's take-home after the car takes its cut, not the number on the offer screen.
Enter those once and you never think about them again. After that, evaluating an order is three numbers off the offer screen and a couple of seconds — faster than typing the post, and far faster than waiting for forty strangers to disagree.
What the Math Still Can't Tell You
Being straight about this: a calculator answers one question well — what does this order pay me per hour after what it costs me to drive it? It doesn't know that the pickup is a restaurant that runs twenty minutes late every time. It doesn't know the drop-off is a gated apartment complex with no visitor parking. It doesn't know it's 9:40pm and this is the last order you'll see all night.
Those things matter, and that's where the groups actually earn their keep. Use the numbers for the math and the drivers for the local knowledge. What you shouldn't do is let someone else's car and someone else's rent decide what your time is worth.
🧮 Get Your Own Answer in Five Seconds
Free, no signup, nothing saved anywhere but your own browser. Set your car up once, then just punch in the offer.
Open the Order Calculator →Frequently Asked Questions
Why do drivers disagree so much about whether an order is worth taking?
Because the offer screenshot only contains half the equation. The payout, mileage and estimated time are visible, but fuel economy, local gas prices and each driver's personal hourly target are not — and those decide the answer. A Prius driver in a slow market and a truck driver in a busy one will correctly reach opposite conclusions about the identical order.
Does gas being more expensive in California actually change whether an order is worth it?
Far less than most people assume on a single order. On a 9-mile delivery, a 50 MPG car in California burns about $0.89 of gas while a 32 MPG car in Tennessee burns about $0.82 — a seven cent difference despite gas costing two dollars more per gallon. Your vehicle's fuel economy outweighs your state on any one order. Across a full year of driving, the state difference does add up to roughly $1,600 for the same car.
Should I subtract gas or the IRS mileage rate when evaluating an order?
Neither on its own. Subtract gas, worked out from your own MPG and pump price, plus a wear figure of roughly 15¢ a mile for an older paid-off car, 22¢ for a newer paid-off one, or 35¢ if you are still making payments. Do not use the IRS rate of 72.5¢ to judge an offer: about a third of it covers insurance, registration and finance charges you pay whether or not you take the order, and it already includes fuel, so subtracting gas on top double-counts the same gallon. Claim the IRS rate at tax time, but decide with the marginal cost.
What hourly rate should I set as my target?
There is no universal number. Base it on your cost of living, what else you could be earning in that hour, and how busy your market is. A useful starting point is your local living wage plus a margin for the wear you are putting on your own vehicle. Set it as take-home after vehicle costs, not as the gross number the app shows, or you will end up chasing a target you have already missed.
How do I figure out my real MPG for this?
Do not use the window sticker. Fill the tank, reset your trip meter, drive a normal shift or two, then divide the trip miles by the gallons it takes to fill back up. City delivery driving with constant stops, idling and short trips typically returns noticeably less than the advertised combined rating, and using the optimistic number makes every order look better than it is.
How should I handle the drive back when the app only shows one direction?
On a single order, assume the return is roughly the same as the delivery distance, since you finish about as far from your zone as you drove. On stacked or batched orders it is often much shorter — a 14-mile triple can end just three miles from the store if the drops chain back toward it. Look at the last drop-off on the map and estimate the return from there rather than doubling the total.
Get the Full GigNova Playbook
Order evaluation is one piece. The free playbook covers real earnings math, mileage tracking, platform tips, and everything the apps won't tell you.
Get the Free Playbook →