My wife pitched the idea in March: instead of guessing which grocery app was cheapest, why not actually track it? For four weeks, I ordered the same rough basket — milk, eggs, bread, chicken thighs, produce, a few pantry staples — through four different delivery or pickup options, and I wrote down every single fee on the receipt. Delivery fee, service fee, small-basket fee, the default tip, and the gap between the in-app item price and what the same item cost on the shelf when I checked in person. It took longer than I expected, and the results were not what I thought they'd be going in.
I'd assumed delivery would just be delivery-fee-expensive and pickup would be basically free. That is not how it played out. The real cost gap wasn't really about the delivery fee line item at all — it was about item markup, and pickup wasn't as free as I'd assumed either.
What I Actually Tracked
Every week I built the same list, ordered it a different way, and logged four numbers: the sticker delivery/service fee total, the default tip the app pre-selected (and whether I had to actively lower it), the total at checkout, and then a spot-check of five items against the shelf price at the physical store later that week. I kept a little notebook on the counter because I knew if I didn't write it down immediately I'd lose track by day two.
The delivery fee itself, on all four options, hovered in a fairly narrow band — a few dollars, sometimes waived above a minimum order size. That part was honest and easy to see. The service fee was where things got murkier. It was usually a percentage of the subtotal, disclosed in small print, and it scaled up as my basket got bigger rather than staying flat. A $60 order and a $120 order had noticeably different service fees even though the effort of picking the order presumably didn't double.
The Tip Default Trap
Here's the thing that actually annoyed me most, more than any fee: three of the four apps defaulted the tip to a percentage of the total order value rather than a flat suggested amount, and it was pre-selected before I even scrolled to that screen. If I hadn't manually tapped in and changed it, I would have tipped several dollars more than I intended on my bigger orders, simply because the tip scaled with grocery prices instead of with effort. Tipping the shopper fairly matters to me and I don't begrudge it — but a default that quietly moves with your order size, rather than staying a flat suggested amount, is a design choice worth noticing before you approve checkout.
Where the Real Money Was Hiding
The delivery fee and service fee combined were rarely more than eight or nine dollars on a typical order. The item markup was the bigger story. When I compared five spot-check items — a gallon of milk, a dozen eggs, a bag of chicken thighs, a box of pasta, and a bag of apples — against the same items' shelf price at the physical store days later, the app prices ran higher on three of the four platforms. Not wildly higher, more like 8 to 15 percent per item on average across my spot checks, but that adds up fast across a full cart, and it's a cost that never shows up as a labeled fee line item anywhere on the receipt. It's just baked into the price you see when you're shopping.
Pickup Wasn't the Free Ride I Expected
I figured pickup — where you order through the app or website and just drive to collect the bags yourself — would sidestep almost all of this. It sidestepped the delivery fee and the tip question, since I wasn't tipping a driver for a curbside handoff the same way. But on two of the four platforms, the item prices for pickup orders were the same inflated in-app prices as delivery, not the in-store shelf price. So I was still paying that markup, just without the fee and without needing to tip. On the other two, pickup pricing matched the shelf price almost exactly, which made it clearly the cheapest option of everything I tested that month.
What Actually Won
Adding it all up over four weeks, the single cheapest way to get the same basket into my kitchen was pickup through the option where item prices matched the shelf. It beat full delivery by a solid margin most weeks, and even beat the other pickup option, because that one still carried the inflated item pricing even without a delivery fee attached.
The most expensive was full delivery through whichever app had the highest service-fee percentage that particular week, combined with a tip default I forgot to adjust down on the first order. That one order alone cost close to eleven dollars more than the same basket picked up in person, once I added everything up afterward.
What Changed for Us
We still use delivery sometimes — when it's genuinely raining sideways or someone's sick, the convenience is worth paying for and I'm not going to pretend otherwise. But for our regular weekly basket, we switched to whichever pickup option matched shelf pricing in our area, and we check the tip default every single time before hitting confirm, regardless of which method we use.
Over the four weeks I tracked this, the gap between my best option and my worst option came out to a little over thirty dollars, just from choosing differently, with the exact same groceries going into the same fridge. That's not because any one app is a scam — it's because the fee structure is built to be skimmed past, not read line by line. Once you actually stop and read every line, the picture changes fast.
If you want to run your own version of this, you don't need four weeks. Order your normal list two different ways in the same week, screenshot both receipts, and compare the total plus five item prices against the shelf. You'll know within one comparison which way your money is actually going.
