How to Evaluate a CJ Product Before You List It

Most “how to dropship” advice stops at the interesting part. You are told to find a winning product, and then handed a checklist that amounts to “look for something with good margin”. That is not a method. This post is the method we built into Volmeris, written out longhand so you can apply it whether or not you use the platform.

The short version: a product is worth listing when it survives four hard gates — margin, net margin, delivery time, and stock — and then wins on a score you can actually reason about.

Start with landed cost, not product cost

The number CJ shows you is the product cost. It is not what the sale costs you. Landed cost is:

landed cost = product cost + shipping cost

Shipping is where most first-time sourcing maths falls apart, because a $4 item with $6 shipping is a $10 item. Two ways to estimate it before you have real orders:

• A flat assumption. Pick one number for the category and apply it everywhere. Crude, fast, and fine when your catalogue is physically homogeneous.
• Weight-scaled. Multiply the product weight by a per-kilo rate. Better when your niche spans a keyring and a play tent.

Volmeris supports both models when it scores candidates, and prefers a live carrier quote over either when one is available. Whichever you use, write the assumption down. An unexamined shipping guess is the single most common reason a “40% margin” product turns out to lose money.

Then work out net margin, because gross margin lies

Gross margin — (retail − product cost) / retail — is the number every sourcing tool shows you, and it ignores every cost that arrives after the sale. The number that decides whether you have a business is net profit per unit:

net = retail − landed cost − processing fees − ad cost − returns allowance

Reasonable starting assumptions, and the defaults Volmeris ships with:

Cost	Default assumption
Payment processing	2.9% of retail + $0.30
Ad cost	20% of retail
Returns allowance	3% of retail

Run a real example. A product costing $8 with $4 of shipping, sold at $25:

• Landed cost: $12.00
• Processing: $25 × 2.9% + $0.30 = $1.03
• Ads: $25 × 20% = $5.00
• Returns allowance: $25 × 3% = $0.75
• Net: $6.23 per unit — 24.9% of retail

That product carries a 68% gross margin and returns 25% net. The gap between those two numbers is the entire post. A product at 45% gross margin with the same assumptions is roughly break-even once ads are paid for, which is why a 45% gross-margin floor is a starting filter and never a decision.

If you sell mostly through paid acquisition, the percentage-of-retail ad model above understates the risk on cheap items — a $12 product does not get cheaper customers than a $40 one. Model ad cost as a flat target cost per acquisition instead, and watch how quickly low-ticket items stop clearing the bar.

Four gates that should reject a product outright

Scoring is for ranking the survivors. These are the gates that come first, and a product failing any one of them is not a close call:

1. Gross margin floor. A default of 45% leaves room for the costs above. Below it, you are working for the supplier.
2. Net margin floor. A default of 15% net. This is the gate that catches the products that look fine on gross and are not.
3. Delivery time ceiling. A default maximum of 15 days, with 10 days treated as ideal. Every day past your customer’s expectation converts into support email and chargebacks, and neither shows up in a margin calculation.
4. Stock depth. A default minimum of 10 units — applied when CJ actually reports a number. Listing something the supplier is about to run out of buys you a cancelled order and a refund; a listing with no reported stock at all skips this gate rather than failing it, so treat those as needing a manual check.

Add one non-negotiable qualitative gate: intellectual-property risk. Anything flagged as branded, licensed, counterfeit, or trademark-adjacent is a hard no, regardless of margin. Volmeris blocks these candidates outright rather than scoring them down, because the expected value of a takedown or a frozen payment processor is worse than any product is good.

Rank what’s left on margin times demand

Once a product clears the gates, the question is priority, not permission. Two signals do most of the work:

• Demand. CJ exposes units sold per listing. It is a rough proxy — it tells you the product moves, not that it moves for a store like yours — but a product with zero recorded sales is a hypothesis, not an opportunity.
• Margin contribution in dollars, not just percent. A 60% margin on a $9 item is $5.40. A 30% margin on a $60 item is $18. Percentage flatters cheap products; your bank account does not care about percentages.

In the Volmeris CJ browser, search results carry estimated margin, units sold, and shipping days where CJ provides them; the Trending tab is pre-sorted so the best margin-times-demand picks surface first. The AI Merchandiser Scout goes further and produces a weighted composite score across net margin percentage, dollar contribution, demand, delivery speed, and risk — we’ll break that scoring down in a later post.

Price the thing deliberately

Sourcing and pricing are the same decision. Before you import, know three prices: the low price you would run in a promotion, the price you plan to list at, and the high price you would test. Compute net margin at all three. If the low price is underwater, you have no room to run a discount — which means you have no room to compete, because eventually someone will.

One small thing that matters more than it should: prices should land on a tidy ending rather than an arithmetic result. A converted price of £20.34 reads like a rounding error. £20.99 reads like a price. Volmeris snaps converted prices to a psychological ending automatically, per unit.

Do the arithmetic before you build the store

The failure mode this method prevents is the expensive one: spending a weekend on branding, theming, and product copy for a catalogue that could never have been profitable. Landed cost, net margin, delivery days, stock, and risk are five numbers you can check in a few minutes per product, before any of that work happens.

If you would rather not check them by hand for every candidate, that is precisely what the AI Merchandiser Scout exists for. Volmeris is free during the public beta — sourcing tools included.