David - MidnightDrop

Co-founder/CTO @ LiztoPro + Amazon FBA seller. Built an AI that finds clean, vetted OA leads — no multipack traps, capped list. Building in public 👇

[David - MidnightDrop @DavidBuildsFBA]

What I'm building for Amazon OA 👇

An AI that scrapes clearance nightly, matches to Amazon via Keepa, and kills pack/size/variant traps before a lead reaches you. Only clean 30%+ ROI flips make the list — capped so deals survive.

Vetted, not scraped. DM for the beta.

[David - MidnightDrop @DavidBuildsFBA]

@ChannelStride Exactly. Rank tells you it sold — not who held the Buy Box or at what margin.

I pull the 90-day price + BSR history side by side. A rank spike that lines up with a coupon run is noise, not demand.

Reading those two apart is where the edge is.

[David - MidnightDrop @DavidBuildsFBA]

I'm a software CTO. I also run Amazon FBA on the side. Watching VAs hand me "deals" that were secretly single units matched to Amazon 3-packs — and cost me money — broke my brain. So I built an AI to kill those traps automatically. Going to share what it finds.

[David - MidnightDrop @DavidBuildsFBA]

Amazon's "virtual multipack" pilot bundles brand ASINs into new multipack listings — and resellers are the ones who'll source a single against them and eat the mismatch.

"Bad analysis"? No. The listing moved under you. That's the trap I'm automating 🌙

[David - MidnightDrop @DavidBuildsFBA]

@1000asins Transparency belongs in the same bucket as gating — a "can you actually move this?" check that has to happen before the PO, not at the FC dock. Expensive lesson to learn at receiving. Good callout 🙏

[David - MidnightDrop @DavidBuildsFBA]

@Josemenchu12 This is the trap in one post. Peak fees, shipping, sales tax — none show up in the gross spread, and all three can flip a "winner" to a loss. The only number that matters is fully-loaded net. If your ROI isn't counting Q4 fees right now, it's lying to you 🙏

[David - MidnightDrop @DavidBuildsFBA]

@fba_realist Agree on the principle. But it's not always poor analysis — Amazon auto-creates virtual multipacks (VMP_ pilot is live now), listings shift after you vet them, and a VA doing 300 leads misses one. The rule's easy; holding it at scale isn't. That's what I'm automating 🙏

[David - MidnightDrop @DavidBuildsFBA]

@fba_realist Point 3 is the underrated one. Sourcing from a valid ungating supplier means a clearance buy pulls double duty — a profitable flip AND the invoice to open the gate on that brand. Most treat sourcing and ungating as separate problems. They're not 🙏

[David - MidnightDrop @DavidBuildsFBA]

@fba_realist This is the post more beginners need to read. When take-home is already under 20%, the costs people forget are what decide profit vs. busywork — shipping, sales tax, and the quietly-wrong "deal" (a single matched to a multipack). Revenue hides all three. Net never lies.

[David - MidnightDrop @DavidBuildsFBA]

Hot take: most "deal lists" are worthless the second they sell to 100 people.

Everyone piles into the same ASIN, the buy box tanks, and your margin's gone before your units land.

A sourcing list should be capped. Scarcity isn't a gimmick — it's survival.

[David - MidnightDrop @DavidBuildsFBA]

@aidenlewiis Points 1-2 are underrated — a retail receipt getting you ungated is basically dead. Which is why for OA/RA the move isn't "win the gate," it's filter commonly-gated brands OUT before you source, so you stop chasing deals you'll never get approved to sell. Solid thread

[David - MidnightDrop @DavidBuildsFBA]

@askjimmysmith The "is Amazon on the listing" flag being inline is the real win. Rank gets people excited, but if Amazon's on it holding the buy box, your share is basically zero. Catching that before you click saves you chasing dead "opportunities." Nice find 🙏

[David - MidnightDrop @DavidBuildsFBA]

@bshah005 Love this — mismatch is the silent killer. One thing I'd add from the sourcing side: by the time it's wrong in the title, the seller already bought the wrong pack. Great copy can't rescue a bad PO. You fix it after the buy, I'm trying to catch it before. Both needed 🙏

[David - MidnightDrop @DavidBuildsFBA]

@tech_lead_bunny This is the exact trap that made me build a validator — a pack count you didn't write, bundled into a "new" ASIN. Sellers find out when the returns roll in. The listing quantity field lies constantly, so you cross-check it, you don't trust it. Good PSA 🙏

[David - MidnightDrop @DavidBuildsFBA]

@FBAwithSuban This should be laminated. Eligibility first saves so much heartbreak. One upgrade: don't just check per-product after you find a "deal" — filter commonly-gated brands out BEFORE you evaluate, so you never fall for a win you can't actually sell.

[David - MidnightDrop @DavidBuildsFBA]

@ChannelStride Exactly — per-unit price is one of my favorites. Catch is sellers game it: a multipack priced near a single to look like a deal, or coupons/Subscribe & Save skewing it. So it votes, it doesn't decide. Stacking the weak signals is the whole game 🙏

[David - MidnightDrop @DavidBuildsFBA]

Building an AI that validates Amazon arbitrage leads:

• scrapes clearance overnight
• matches to Amazon via Keepa
• checks brand, pack, size + variant before I see it
• keeps only 30%+ ROI flips

No multipack traps. No variant mismatches. Just clean math.

[David - MidnightDrop @DavidBuildsFBA]

@FbaAbe Solid framework. One flag on the sourcing-VA route — it's the hire where mistakes cost most. A single unit matched to a multipack wipes out five good flips. Delegate the finding, sure, but never the verifying: keep a hard pack/variant check before every buy.

[David - MidnightDrop @DavidBuildsFBA]

@Nuurastros Mostly agree — "automated my whole business with AI" is usually slop. One exception though: narrow checking. AI won't magically source for you, but verifying pack/size/variant across thousands of listings is just the "tedious admin" you mentioned — applied to sourcing.

[David - MidnightDrop @DavidBuildsFBA]

@askjimmysmith Brutal lesson, great catch. The deeper issue: those leads were never yours alone. Same mechanism whether it's a double-dealing VA or a cheap list sold to 100 buyers — the moment identical leads hit many sellers, the buy box races to the bottom. Sourcing only survives capped.

[David - MidnightDrop @DavidBuildsFBA]

@MonikaFlips So true — instructions can't cover every edge case. The one place I'd push further: sourcing verification. A VA matching a single unit to a multipack isn't a judgment miss, it's a checking miss — thousands of tiny comparisons where humans tire. That one task I ended up automating