Everyone obsesses over finding a "winning" product. Almost nobody obsesses over landed cost drift — and it's the actual silent killer of margins at scale.

Here's what happens: you find a product, supplier quotes $4.20/unit, shipping looks like $2.80, you price at $29.99 and the math looks great on a spreadsheet. Three weeks in, you're running 200 orders/day. Suddenly:

• Your supplier's "stable" price moves because you're now a bigger order, not a sample buyer

• Shipping zones you didn't account for start eating $1-3 extra per order

• Return/refund rate (which nobody models at the research stage) quietly takes another 4-6% off net

The product didn't stop winning. Your unit economics did — and most people never notice until the account is already bleeding.

The fix isn't a smarter ad. It's re-running your cost math at three volume tiers (10/day, 50/day, 200/day) using actual supplier quotes at those tiers, not the sample price, before you scale spend. If the margin doesn't survive tier 3, you don't have a winning product — you have a working demo.

Vet the supplier relationship at scale, not at sample size. That's the whole game.

Ran research for dropshippers for a while now, and the pattern is always the same: someone finds a "winning product" on a free TikTok compilation or a Telegram group, launches it, and wonders why their CPA is $40 on a $25 product.

Here's the thing nobody tells you: by the time a product hits a public "winning products" list, it's already been running ads for 2-3 months. You're not finding a winner — you're finding the tail end of someone else's winner, right as saturation kicks in and CPMs spike.

A few things that actually separate a live opportunity from a dead one:

1. Ad library recency, not just ad library presence. Anyone can screenshot an ad library and say "look, this product has ads running." What matters is how many new advertisers started running it in the last 7-14 days. Rising new-advertiser count = still early. Flat or declining = you're late.

1. Search trend shape, not search trend existence. A flat, high search volume for months means it's a stable niche product, not a spike opportunity — different game, different margin expectations. A sharp recent slope up is the actual signal.

1. Landed cost drift. Suppliers change pricing constantly based on order volume from other dropshippers. The margin math you did last month on a product is not the margin math today. If you're not re-checking landed cost before you scale spend, you're guessing.

1. Supplier reliability is the silent killer. More stores die from a supplier ghosting mid-fulfillment than from picking a "bad" product. Vet the supplier before you vet the product idea.

None of this guarantees a winner — execution, creative, and offer still do most of the work. But chasing stale data with good execution still loses to fresh data with mediocre execution, more often than people want to admit.

Curious what others here are seeing on saturation right now — anyone tracking a category that's turning over faster than usual?