Last Tuesday I watched a 52-item Target order sit on the open metro for eleven minutes. $14 base. Estimated tip was $0. Customer lived way out, like 18 minutes from the store with no other deliveries in that direction.
Somebody grabbed it. Whoever that was spent the next two hours shopping and driving for what worked out to almost nothing after vehicle costs. I know because I ran the numbers out of morbid curiosity while I sat in the Meijer lot waiting for something better.
A year ago that would've been me.
The math nobody wants to do
Shipt's V2 pay algorithm is a black box. Used to be $5 plus 7.5% of the order. Now it's "effort-based." What does that mean? Nobody outside Shipt knows and they're not telling.
They guarantee $16/hour minimum. Based on their time estimates though, not yours. And those estimates don't touch your actual costs.
Gas is the one everybody thinks about. But gas is maybe a third of what it costs to run your car. Tires, oil changes, brakes, insurance, depreciation. Your car loses value every single day you do this job. The IRS puts vehicle cost at 72.5 cents a mile for 2026. That number accounts for all of it.
Then there's the drive back.
You deliver the groceries. Great. Now you're sitting in some subdivision 18 minutes from the store. Gotta drive all the way back, or at least most of the way, to get your next order. That return trip? Unpaid. Those miles? Unreimbursed. Your tires still wore down and your gas tank is still lighter and you burned another 15 or 18 minutes getting back to where the orders are.
So let's do the actual math on that $14 order. Over an hour shopping a 52-item Target run, finding stuff across departments, texting the customer about substitutions, waiting in the checkout line. Call it 65 minutes in the store. Eighteen to the customer. Another 18 back because you're going the same distance in reverse. That's about an hour forty. Round trip is around 22 miles. At 72.5 cents a mile that's roughly $16 in vehicle operating costs.
$14 minus $16. You lost money. Actually lost money. Paid for the privilege of shopping someone's groceries.
Even if we're generous and say you caught another order near the drop-off and only drove 10 minutes back, you're still at an hour and a half total and maybe $5 in your pocket.
Same order with a $12 tip? Around $10 net for the same work. Not great but you can live on that. Without the tip you're literally paying to work.
Why I stopped taking them
I don't have a catchy motto about it. I just started paying attention.
Tracked every delivery for about four months. Address, order total, what Shipt paid, whether they tipped, how much, how long it took. Put it all in a spreadsheet at first, then switched to something less painful later on.
Patterns showed up fast. The addresses that stiffed me in January stiffed me in February. March. April. Not a single one of them flipped. I had this one house, a big house, consistently $150+ orders. Delivered there three times thinking they'd come around. $0. $0. $0. I was subsidizing this person's grocery delivery with my own gas money. That's what was actually happening.
Most shoppers who've been around figure this out eventually. That's where DND lists come from. Do Not Deliver. Some people keep them in their notes app. Some label the address in Google Maps with "NT" or "no tip" or "DND" so it shows up next time they're near the area. Some just remember. But everybody who sticks with Shipt long enough has addresses they won't go back to.
Acceptance rate used to be a thing. Keep it above 80% or Shipt buries you in the queue, that was the fear. Scared me too, my first couple months. Took all kinds of garbage to keep the number up. Shipt got rid of it though. Your star rating and preferred shopper status are what drive order flow now. So there's no penalty for passing on a bad offer.
There's also this argument that taking no-tip orders enables the behavior. If every shopper refused them, customers would have to start tipping. I get it. Makes sense in theory. In practice some people just don't tip and nothing's gonna change that. All I can control is whether I'm the one who shows up.
When I still take them
I'm not rigid about it. Situations matter.
Brand new customer with a small order, close to the store? Sure. I've had first-timers hand me a $20 at the door because they didn't know how tipping worked in the app. Had another one add $15 two hours later. You can't judge new members by a $0 estimate, that number is meaningless when there's no order history behind it.
Quick runs barely register. Four items, half a mile. In and out in ten minutes, burned basically no gas, put a negligible number of miles on the car. Even with zero tip I didn't lose money, I just didn't make much. There's a difference.
Slow days change my thinking completely. I sat in a parking lot for over an hour last month on a Wednesday afternoon. Nothing. Eventually a $10 no-tip order came through, close to the store, 12 items. Took it. Made about $7 net for 25 minutes of work. Not good, but $7 beats $0, and I was losing my mind sitting there.
Then there's the Preferred Shopper play. This one's real. Deliver to someone, give good service, they add you as preferred. Now you get their orders before they hit the open metro. And I've noticed, not every time but enough times to pay attention, that customers who start off not tipping sometimes start once you become "their" shopper. Something about the relationship being personal changes things. Worth keeping in mind for new addresses you haven't seen before.
The estimated tip is mostly fiction
Shipt puts an estimated tip on the offer screen. Algorithm pulls from the customer's history, order size, trends, whatever else. Sometimes it's right.
Lot of times it's not.
I've had $0 estimates turn into $18. Had a $14 estimate turn into nothing. Had estimates that were close. Had estimates that were off by $20 in both directions. For new members with zero order history the estimate is generated from, I don't know, vibes? It doesn't track to anything real.
Which brings up the actual problem. Every shopper eventually realizes they need their own records. Because the app's estimate isn't reliable enough to bet your hourly rate on.
So you do what everyone does. Open Google Maps, find the address, add a label. "Good tipper." "NT." "Cash $15." "Stiff, never again." Little text tags that show up when you search the address or happen to scroll past it on the map. Works okay until you forget to label one because you were rushing to your next order. After a couple weeks you've got labels on half your deliveries and gaps on the rest. I tried a spreadsheet for a while too, matching tip notifications to deliveries from three days earlier. Cross-referencing screenshots. Got old real fast.
The question "should I take this no-tip order?" is really a different question: have I delivered here before, and what happened? Two seconds to answer if you know. Ten minutes of agonizing and guessing if you don't. The knowledge is what matters, not the rule.
What actually changed things for me
Dropped the blanket rules. No more "I always decline under $X" or "I never take no-tip orders." Too rigid. Doesn't account for slow days, new customers, short runs.
What works is knowing the addresses.
I track everything now. Every delivery, every tip, every stiff. Been doing it long enough that most repeat addresses in my zone have a history I can check. Four orders, zero tips? Not a chance. Tips $12 every time? I'll grab that even if base pay is low because I already know what's coming on the back end.
New address with no history, I'll take it once if the base pay covers my floor. Anything below $17/hour net after vehicle costs and I'm not interested. But the first delivery to an unknown address is an investment. Either they tip and I've found a new reliable stop, or they don't and I've added them to the DND list. Both outcomes are useful.
The preferred list is the long game. Put your best work into customers who treat you right. Tip well, rate you five stars, don't change half the order mid-shop. Over months your schedule fills with their orders and the open metro becomes a backup instead of your main source.
I still take an occasional no-tip order. But I don't sit in the lot agonizing anymore. I just check the address and I already know.
Auto Tip Map is an Android app that tracks your orders, tips, mileage, earnings stats, and estimated fuel costs for Shipt shoppers. When that address pops up on the offer screen again, you'll already have the answer.