Five causes, roughly in order of how many stores they take out. Every one of them is avoidable, which is the good news.
01
The generic store
Importing the same AliExpress products as thousands of other people, with zero differentiation. No niche, no point of view, just a grid of unrelated items. The moment a customer sees your $39 gadget, they open another tab — and Amazon has it for $22 with next-day delivery, or Temu has it for $9. You've built a worse version of a store that already exists, and priced it higher.
Instead Pick one niche and build a store that feels like a destination for that one thing. You are not competing on price with Amazon and Temu — that fight is unwinnable. You're competing on being the place that specific person actually wants to buy this specific thing from.
02
Underestimating what ads cost
The "just start with $5/day" advice is dead and has been for years — Advertising costs vary by audience, season and campaign; there is no fixed budget that guarantees useful results. At $5/day you're not testing, you're buying a rounding error of data and then making decisions on it. Meanwhile product tests can fail, so people spend a little, lose it, conclude "ads don't work," and stop right before the point where the numbers start meaning something.
Instead Budget $30–50/day per product tested, and count on most of those tests losing money. Your first few hundred dollars of ad spend is tuition, not profit. If you can't fund that comfortably, use the organic path in §F instead — but don't run the paid path on a budget that can't produce a signal.
03
Shipping that destroys trust
Unvetted China-only suppliers with 15–30 day delivery windows. The customer forgets they ordered, then panics, then emails you angry, then requests a refund, then files a chargeback and leaves a review. Every one of those costs you the sale, the ad spend that won it, and a piece of your payment-processor standing.
Instead Vet suppliers before you sell anything, and state real shipping times on the site. Cheap China sourcing is fine for testing, but the moment a product proves itself, move it to a supplier with US warehouses and 2–7 day delivery — even if it costs you margin. Speed is retention.
04
The wrong niche
Low-ticket commodity items with razor-thin margins. A $12 phone accessory has to clear product cost, shipping, payment fees, and ad spend out of a couple of dollars — the math simply doesn't close, no matter how good the ad is. People pick these because they're cheap to test, which is exactly the wrong reason.
Instead Choose on margin headroom and perceived value, not on how cheap it is to try. If the product can't support a sell price of 3× its landed cost, it can't support a business. There's a calculator in step 4 of the checklist — use it before you fall in love with anything.
05
Treating it as a lottery ticket
No numbers tracked, no idea which product is actually profitable after ads and refunds, no email list, no repeat customers. Every sale has to be bought fresh from Meta at full price, forever. Stores in this state can look busy and still be quietly losing money on every order — the owner usually finds out months late.
Instead Track per-order margin from your very first sale, and start collecting emails on day one. Repeat buyers are what make the ad math work: the first purchase can break even if the second and third are nearly free.