Neither one shows up on your dashboard. Both are sitting in your data right now — and you can find them yourself in an afternoon.
A growing store can be bleeding money and still look healthy. Revenue is up. Ads are profitable. You’re busy. Nothing screams.
That’s because the money rarely leaves through the front door, where you’d see it. It leaves through the seams — the gaps between the tools you actually check. And once you’ve taken apart enough stores, you see the leaks aren’t random. They come in exactly two kinds.
Most stores have both. Here’s each one, shown through a real store — so you can go find yours.
Estimate my leak — 2 minutes →
Want the shortcut before reading on? Plug in your numbers now.
The first store did about $2M a year selling pet supplements. Call it Northpaw. Its marketing worked fine — that was never the problem. The ads made money. The traffic was real. The buyers were ready. They just kept hitting walls nobody knew were there. Three of them.
Their credit-card processor ran an ID check. Before a card would go through, the customer had to upload a government ID — a driver’s license or passport. To buy pet supplements. Most people did what you’d do. They left.
Card orders died at 3.5x the normal rate. Here’s the part that proves they wanted to buy: 439 customers refused to quit. They found another way to pay and bought anyway. Another 685 just left, and took about $138K with them. Those weren’t cold leads. They had the product in the cart and a working card in their hand.
A wall where you think you have a door.
Tap an ad inside Instagram and the link doesn’t open in Safari or Chrome. It opens in Meta’s own built-in browser. That browser blocks cookies, breaks autofill, and breaks payments. The customer never picked it and never noticed. But it cut their conversion rate in half — then in half again.
The fix is one banner, and you can hand it to a developer today. The browser can be detected from its User Agent string (it contains FBAN, FBAV, or Instagram). When you spot it, show one line on the cart and checkout pages only: “For the best checkout, tap ⋯ and open in your browser.”
One tap sends the customer to Safari or Chrome, where they convert about 7x higher. One or two days of work. A few hundred dollars, once. It pays for itself the first week and never stops.
Their abandoned-cart emails were bringing back about $5,000 a year. The reason was one setting. The emails fired on “Started Checkout” — so they only reached people who’d made it to the final page. Everyone who dropped off earlier, at the cart, got nothing. That’s the warmest audience a store has, and the recovery emails couldn’t see them.
| Setting | Orders | Revenue |
|---|---|---|
| “Started Checkout” (theirs) | ~30 | $5,000 |
| “Added to Cart” @ 3% | ~374 | $61,781 |
| “Added to Cart” @ 5% | ~624 | $116,820 |
None of these three were marketing failures. The ads did their job. They brought ready buyers to a store that kept turning them away at the door. That’s the signature of Way One: the loss always happens after the click — in a place the media buyer doesn’t own and the developer was never told about.
Estimate my leak — 2 minutes →
Curious if Way One is costing you money too? Plug in your own numbers.
The second store sold kids’ clothes and did about $1.1M a year. Call it TinyTrackers. It had the opposite problem. Nothing was broken. Checkout worked. Ads were profitable at 2.94 ROAS. People bought. They just bought small — one item, over and over — and nobody had built anything to ask them for more.
Here’s the math that should stop you. You can run the exact same math on your own store tonight:
Here’s what makes this Way Two and not just “low AOV.” Their best campaign ran one offer — “15% off orders over $250” — and it returned a ROAS of 4.14 against a blended 2.94. Forty percent better than everything else in the account.
The customer clicked that ad. They landed on the site ready to hit $250 and earn the discount. And the site said nothing. No banner. No progress bar. No “add $40 more to unlock 15% off.” The most profitable offer in the whole business lived only in the ad — because the space between the ad and the site belongs to no one.
The cart progress bar is the highest-leverage thing most stores skip — and it’s a built-in feature on every major platform. Set a threshold just above your two-item order value. Then show a live bar on the cart: “You’re $34 away from 15% off — add one more.”
It turns the discount into something the customer unlocks instead of something you give away. And it turns your most-visited page into a salesperson that never gets tired. The proof it works is already in this store’s ad account. It just never made it onto the site.
Way Two is quieter than Way One, and meaner for it. Nothing looks wrong. Every order is a small win. The dashboard is green. The missing money never shows up as a loss — because you can’t see a second item that was never added. You only see the single-item order, sitting there looking perfectly fine.
Estimate my leak — 2 minutes →
Not sure how much Way Two is costing you? See the estimate.
You don’t need us to run this. You need an hour with your own numbers and the nerve to look. Seven checks, split by the two ways.
Write down what you find. Price each gap against your own AOV and margin. Some of these you can fix yourself this week — the banner, the trigger, the progress bar are all in this article, and they’re yours to take. Others go deeper, and are worth doing right.
But the fix list isn’t the point. The point is where you look. Stop watching the front door — revenue, ROAS, the numbers that are already green — and start watching the seams. Once you see it, you can’t un-see it. The money was never hiding. It was just standing somewhere no dashboard was pointed.
That’s the whole article. No form, no download, no catch.
We wrote it because these two patterns are in almost every store we open, and most founders have never had them named.
Running the seven checks by hand is the honest way to do it, and we’d rather you did. But if you’d like the shortcut — six numbers you already have on your dashboard, and a rough estimate of what each leak is costing you per year — we built a small tool that does the arithmetic for you.
Estimate my leak — 2 minutes →
Start with the afternoon if you’d rather — most of what’s leaking, you can find on your own.