Can't tell why your sales dropped? Sledge reads your store data and ranks the leaks by what fixing them is worth. Free for 14 days →

// Level 02 Getting Sales · 2.2

Illustrated violet mountain under a wide halo of light, with a golden path leading toward it.

Paid ads without the bleed

Updated August 2026 9 min read
THE ROAS YOU NEED, BEFORE YOU WRITE AN AD HARD TO WIN 5x 4x 3x 2x 1x 20% 30% 40% 50% 60% 70% 80% 90% MEDIAN DROPSHIPPING PRODUCT: 77% → 1.30x 20% margin → 5.00x needed YOUR GROSS MARGIN NOTHING IN THE AD ACCOUNT MOVES THIS LINE.

Our own calculation, plotting break-even ROAS against gross margin. The median figure is from a study of 228 dropshipping products.

The short answer

Before you spend anything, work out your break-even ROAS: divide 1 by your gross margin. At 50% margin you need 2x just to break even. At 25% you need 4x, which almost nothing achieves on cold traffic. Budget roughly 50 conversions' worth of your target cost per order for a real test, kill an ad set at about 1.5 times break-even cost per order after $100 to $150, and scale winners by 20 to 30% a day rather than doubling.

One number decides whether ads can work at all

The useful work happens before you open your ad account, and it takes about thirty seconds.

BREAK-EVEN ROAS = 1 ÷ GROSS MARGIN
A $40 product costing you $18 has a 55% margin. 1 ÷ 0.55 = 1.82x

That is the return you need just to stop losing money. Not to profit. To break even.

Here is what it looks like across the range, with the median row taken from a study of 228 dropshipping products:

Gross marginBreak-even ROASRealistic?
20%5.00xAlmost never on cold traffic
25%4.00xVery rare
30%3.33xHard, needs strong creative
40%2.50xPossible, tight
50%2.00xWorkable
60%1.67xComfortable
77%1.30xMedian dropshipping product
85%1.18xLots of room to test

The same study found products under 30% margin averaged a break-even ROAS of 7.93x. That is not a difficult target. That is a product that cannot be advertised profitably, no matter how good the ads are.

Your break-even ROAS was decided the day you picked the product. The ads just reveal it. The Dropshipping Playbook

This is why chapter 1.3 put "the maths never worked" above every other failure. A bad ad costs you a test budget. A bad margin costs you every test budget you will ever spend.

What you can pay per order

Break-even ROAS tells you whether ads are viable. Cost per acquisition tells you what to bid. It comes from the same numbers.

Selling priceMax CPA at 3x ROASWhat that buys you
$25$8.33Roughly one thousand impressions. One.
$40$13.33Workable with sharp creative
$60$20.00Room to test angles
$100$33.33Room to retarget properly
$150$50.00Can outbid most competitors

Your entire acquisition budget for a $25 order buys about a thousand impressions. If your product needs more than one thousand people to see it to produce a sale, and it almost certainly does, the maths has already lost.

This is the real reason experienced dropshippers drift upward in price. Not greed. Headroom.

The $2,000 testing myth

Almost every course teaches the same thing: budget two thousand dollars or more for the testing phase before you can expect to find a winner. One merchant put the obvious question to the Shopify Community.

Merchant report

A merchant explained that every Facebook ads course they had taken said the same thing, that testing requires $2,000 or more to find profitable ad sets. They proposed a rule instead: turn the campaign off after $500 of spend with no add-to-carts, rather than spending $2,000 to discover there were no sales.

The reply from a certified Meta ads strategist was blunt. Ad success does not depend on how much you spend. It depends on the audience, the purpose, and the quality of the campaign. They did not believe the $2,000 rule.

Shopify Community, "Turning off a campaign after spending $500 and no add to carts"

The merchant's instinct was right, and their number was still too high. Zero add-to-carts after $500 is not a marginal result you need more data to interpret. It is an answer, and it arrived several hundred dollars earlier.

Kill rules, written down before you start

The point of writing these down in advance is that you will not want to follow them later. Every losing campaign feels like it is one more day from turning around.

$100 to $150 spent
Zero add-to-carts
Kill it. Nobody wants this at this price, from this creative, shown to these people. More budget buys the same answer.
CPA above 1.5x
your break-even CPA
Kill it. This is the standard threshold. It is far enough above break-even to be a real signal rather than noise.
Add-to-carts healthy
No purchases
Do not kill the ad. The store is the problem, not the traffic. People want it and stop at the cart. Level 03 covers that.
Profitable at small spend
Scale by 20 to 30% a day. Doubling overnight resets the algorithm's learning and usually costs you the winner you just found.
Under 500 site sessions
You cannot read your conversion rate yet, so you cannot separate a bad ad from a normal store. See chapter 2.1.

The third row is the one people get wrong most often. Healthy add-to-carts with no purchases means the ad did its job. Turning it off and blaming the creative is how a store problem gets misdiagnosed as an ad problem for months.

What to budget

Two different questions get confused here: what you spend per day, and what a full test costs.

TEST BUDGET = TARGET CPA × 50
A $40 product at 3x ROAS: $13.33 × 50 ≈ $667 to properly test it

Fifty conversions is roughly where results stop being noise. That figure is per product, not per store, which is why testing five products at once is a five-figure exercise and not a beginner's plan.

$30 to 50Per day, bootstrap stage, on one or two products
30/60/10Testing, scaling, retargeting split past $3k a month
$5 to 10Per day floor, enough for two weeks of signal

If those numbers are out of reach, that is worth knowing now rather than in week three. It does not mean stop. It means paid traffic is not your first channel, and chapter 2.3 is where you should be.

Where the money leaks

Separate from budget, these are the recurring ways spend disappears without producing information.

That last one is the quiet killer. Without a break-even number, a 1.8x ROAS looks like a mediocre result you might improve. With one, you know instantly whether it is a winner to scale or a product to drop.

The one lever that is not in your ad account

You cannot lower CPM. Meta CPMs are up around 89% since 2020 and you are bidding against better funded advertisers than yourself.

What you can change is what a customer is worth when they arrive.

Raising average order value does not change your break-even ROAS, since that is set by margin. It raises the cost per order you can afford, which is the number your ad account actually competes on. A second item in the same order does not pay a second acquisition cost.

A $40 order at 55% margin gives you $22 of gross profit to spend against. Turn that into a $62 order at the same margin and you have $34. Same ads, same CPM, same audience, and suddenly you can outbid the people you were losing to.

That is Level 04.

Common questions

What ROAS should I aim for?
Aim for your break-even ROAS plus enough margin to cover fees, refunds and your own time. Most sellers target 3x to 5x depending on their margin. But "3x is good" is meaningless without your margin: 3x on a 25% margin product still loses money, while 1.5x on an 80% margin product is profitable.
How long should I let a campaign run before judging it?
Give it enough spend rather than enough days. The kill thresholds on this page are spend-based for that reason. On Google Ads specifically, community guidance suggests planning for at least six weeks before drawing conclusions, and budgeting as though you will run for six months.
Can I start with $5 a day?
You can get directional signal. You cannot run a proper 50-conversion test. At $5 a day, treat it as learning which creative gets attention, not as validating a product. And be honest that the two-week result will not be conclusive.
My ROAS looks fine but I have no money. What's wrong?
ROAS is revenue over ad spend, so it ignores product cost, fees and refunds. A 3x ROAS on a product with 25% margin is a loss. Track contribution margin per order instead, and remember that a payment reserve can hold a share of that revenue for months, which chapter 1.3 covers.
Illustrated violet dunes with a yellow sun low over the ridge.

One number pays for the rest

What a single order is worth decides what every other fix can afford. Chapter 4.1 is where that number moves.

Go to chapter 4.1