What profit margin do I need for dropshipping?

The common working rule is a ~3x markup over your landed cost (product + shipping), because paid traffic will eat a third or more of your price. At 2x, one bad ad week puts you underwater; at 3x you have room for the CPA, fees, refunds and still a real margin. Cheap products can need higher multiples because fixed costs weigh more.

How do I check demand for a product?

Triangulate three free signals: search volume for the product's terms (tens of thousands of monthly searches suggests a real market), Google Trends over five years (steady or growing beats spiking), and competitor evidence — stores visibly selling it plus ads that have run for months in Meta's public Ad Library. One signal can mislead; three rarely do.

Should I order a sample before selling a dropshipping product?

Yes, always, before scaling. Photos lie, supplier specs lie, and your refund rate is set by the gap between the product page and the box the customer opens. A sample is the cheapest insurance in e-commerce — the checks in this guide tell you whether the product is worth the sample, and the sample tells you whether it's worth your money.

Field guide

How to validate a dropshipping product — the go/no-go checklist (2026)

Q: How do I know if a dropshipping product will sell?

You never know — you estimate, then test small. The estimate comes from six checks: the product solves a visible problem; the numbers support roughly a 3x markup over landed cost; it's light and compact to ship; demand is year-round rather than a seasonal spike; at least one competitor is demonstrably selling it at scale; and it has natural upsells. A product failing two or more checks is usually a no-go.

Most product failures were visible before the first ad dollar. Six checks separate a candidate worth testing from an expensive lesson — here they are, with the numbers.

Updated 19 August 2026 · by the team behind Savva HQ · ~9 min read

Honesty first: validation doesn't guarantee winners — it filters obvious losers. That's worth a lot: most failed products fail for reasons this checklist catches in an afternoon, before you've paid for ads, apps and a month of hope. Run every candidate through all six; be suspicious of your own excitement on any product that fails two.

Check 1 — Does it solve a visible problem?

Products that fix something you can show (posture, cable chaos, a cleaning chore, pet hair) sell through demonstration — and demonstration is what short-video traffic runs on. “Nice to have” products can win, but they need brand and audience, which a new store doesn't have. First stores should bias hard toward problem-solvers.

Test: can you describe the before/after in one sentence a stranger instantly gets? If explaining takes a paragraph, ads will be expensive.

Check 2 — Does the margin math survive?

The working rule: target roughly a 3x markup over landed cost (product + shipping to customer), because paid traffic typically eats a third of the price:

price ≈ 3 × (product cost + shipping) — then sanity-check against the market

At 2x, one bad ad week puts you underwater. Run the candidate through the full seven-line formula in the real-profit guide with a realistic CPA before you fall in love. Very cheap products (<$20 retail) often need higher multiples — fixed costs weigh more per order.

Check 3 — Is it light, compact and unbreakable?

Boring check, kills the most products in practice. Heavy or bulky = shipping erases margin. Fragile = refunds and 1-star reviews. Electronics with batteries = shipping restrictions and defect rates. The physics of the parcel is part of the product.

Check 4 — Is demand evergreen?

Open Google Trends, set 5 years. You want steady or gently growing interest — not a single viral spike from eight months ago (you're late) and not a hard seasonal sawtooth (unless you're deliberately playing seasons). A trend that already died once will die on you too.

Check 5 — Is someone already selling it at scale?

Counterintuitive for beginners: a competitor selling this product successfully is a green flag, not a red one. It proves demand and a workable funnel. What you're looking for:

A store visibly selling it (check their best-sellers — the competitor research guide shows how, from public data).
Their ads running for months in Meta's public Ad Library — long-running ads are paying ads.
Room to differentiate: a real angle they're missing (bundle, audience, objection their page ignores — read their 1-star reviews for it).

Zero competitors usually means zero demand — or a graveyard you can't see. Ten identical stores with aggressive ads means you're late to a knife fight. The sweet spot is proven demand + a visible gap.

Check 6 — Does it have natural upsells?

Your ad buys the customer, not the sale — profit usually lives in order value. A product with obvious companions (refills, cases, a 2-pack, an accessory) can carry a real average order value. A strict one-and-done product must win on its own margin, which is rarer.

Then: triangulate demand with numbers

Search volume: tens of thousands of monthly searches for the product's terms suggests a real market (free keyword tools give ballparks — ballparks are enough here).
Trends (5y): steady or growing, per Check 4.
Social proof of virality: on product videos, shares matter more than likes — a video shared half as often as it's liked is being sent to someone, which is purchase-adjacent behavior.

Any single signal can mislead. Three pointing the same way rarely do.

📌 The verdict framework we use: GO — passes all six, demand triangulated → order a sample and test small. RESEARCH — fails one check or a signal is unclear → dig on that specific point, don't launch on hope. NO-GO — fails two or more → drop it without regret; the next candidate is free.

The two steps validation can't replace

The sample. Order one before scaling, always. Your refund rate is set by the gap between the product page and the box that arrives. No checklist sees inside the box.
The small test. Validation earns a product a test budget, not a rollout. Real customers voting with money is the only check that's final.

The honest bottom line: product research fails in two opposite ways — launching on excitement without checks, or researching forever without launching. Six checks, three demand signals, one sample, one small test. That's a complete, finite process. Run it, decide, move.

Run this checklist on 50 products at once

Savva's research scans real stores and live signals, scores candidates on these exact criteria, and returns a shortlist with a GO / RESEARCH / NO-GO verdict on each — honestly marked when a signal (like evergreen) can't be verified. You review the shortlist, order the samples.

Try the research free

See what else Savva does day to day →

Common questions

How do I know if a dropshipping product will sell?

You estimate, then test small. Six checks: visible problem, ~3x margin math, light/compact, evergreen demand, a competitor proving it at scale, natural upsells. Failing two or more = usually no-go.

What profit margin do I need?

The working rule is ~3x markup over landed cost, because paid traffic eats about a third of the price. At 2x one bad week puts you underwater; cheap products often need more.

How do I check demand?

Triangulate: search volume (tens of thousands/month), Google Trends over 5 years (steady beats spiking), and competitor evidence — visible sales plus ads running for months in the public Ad Library.

Should I order a sample first?

Yes, always, before scaling. The checklist tells you if the product deserves a sample; the sample tells you if it deserves your money.