What should an online store measure before it scales ad spend?

Most online stores that lose money on ads don't lose it on a bad week. They lose it the month after they double the budget, because the numbers they scaled on were never telling them what they thought. The platform said the campaign was profitable, the bank account disagreed, and nobody could say which one was wrong.

Before you scale ad spend, you need four things you can trust: the ROAS your margin actually needs, purchase tracking that matches your store's own orders, a clear split between new customers and people who'd have bought anyway, and enough conversion data for the bidding to work with. Get those right and scaling is a controlled test. Skip them and it's a guess with a bigger budget.

Why scaling ad spend on the wrong numbers loses money

Ad platforms report the sales they can attribute to themselves. That's useful, but it isn't the same as profit, and it isn't the same as the sales your ads caused. Three gaps open up as spend grows:

• Revenue isn't margin. A campaign returning $3 for every $1 looks healthy until you remember the product, the shipping and the payment fees came out of that $3 first.
• Attributed isn't incremental. Some of the purchases a platform claims would have happened without the ad: returning customers, people who searched your brand name, people who were already on their way to checkout.
• Two platforms can claim one sale. Google Ads and Meta each count the purchases they touched. Add their reports together and you'll usually get more sales than your store actually took.

None of this matters much at a small budget, because the losses are small. It matters a great deal at the point you decide to spend more, because every gap scales with the budget.

Your break-even ROAS: the number every other ecommerce metric is judged against

Return on ad spend (ROAS) is revenue divided by ad spend. Your break-even ROAS is the point where the revenue from ads only just covers what it cost to make, ship and sell those orders. Below it, you lose money on every sale before you've paid rent or wages.

How to work out break-even ROAS from your gross margin

Shopify publishes the formula in its guide to break-even ROAS: break-even ROAS is 1 divided by your gross profit margin, where the margin is revenue minus variable costs (product, shipping and fees), divided by revenue. Its worked example is a $100 product with $60 of variable costs. That's a 40% margin, so the break-even ROAS is 2.5: you need $2.50 back for every $1 of ad spend just to stand still.

1. Take your average order value from your store's own reports, not the ad platform's.
2. Subtract the variable costs of that order: cost of goods, shipping you pay, packaging, payment and platform fees, and the cost of the returns you typically take back.
3. Divide what's left by the order value to get your gross margin.
4. Divide 1 by that margin. That's the ROAS you need just to break even, and any target ROAS you set should sit above it.

Why platform ROAS isn't the same as profit

The chart is why a single "good ROAS" doesn't exist. A ROAS of 3 is comfortable on a 50% margin and loses money on a 30% one, where you'd need 3.33 to break even. Work out your own number per product range if your margins differ, because a sale on a low-margin bestseller and a sale on a high-margin accessory are worth very different amounts to you, even when the platform reports them as the same.

Is your purchase tracking accurate enough to scale on?

Smart Bidding spends your budget on the purchases it can see. If it's seeing the wrong ones, more budget just buys more of the wrong thing faster. Check three things before you raise spend.

Reconcile platform purchases with your store's orders

Pull a month of orders from your store and compare the count and revenue with what Google Ads, Meta and GA4 report for the same month. They won't match exactly, and they shouldn't: each platform only sees what it can attribute. What you're looking for is a gap that doesn't make sense, such as a platform reporting more purchases than the store took, or purchase values that ignore discounts and refunds.

That's how we report on On The Nose Co's ad results: platform-attributed purchases are assessed separately from Shopify sales, so neither number is mistaken for the other.

Count each purchase once

• Browser and server events. If you send Meta purchases from both the Pixel and the Conversions API, Meta's deduplication documentation says the Pixel's event ID must match the Conversions API's event_id, and the event names must match too. If they don't, one sale can be counted twice.
• Refunds. Google's GA4 ecommerce guide measures refunds with a refund event carrying the original transaction_id. Without it, GA4 keeps every refunded order as revenue.
• Lost signal. Google says enhanced conversions can improve the accuracy of conversion measurement by sending hashed first-party data, such as the email address a customer checks out with.

How to separate new customers from repeat buyers and brand searches

The cheapest sales in any account are the ones your ads didn't really make: repeat customers and people searching your brand name. They make ROAS look excellent, and they're the reason a campaign that looked brilliant at $50 a day goes flat at $200 a day. Once those easy sales are bought, extra spend has to find people who've never heard of you, and they cost more.

Measure new customer sales on their own

Google Ads has customer lifecycle goals for this. Its New Customer Value mode tells Smart Bidding to bid higher for new customers than existing ones on Search, Performance Max, Shopping and Demand Gen, and it needs at least one purchase conversion goal. Even if you don't use the setting, the split it relies on is the one to report: what share of your ad revenue came from people who'd never bought before?

When we shifted Scotch & Soda's paid social towards prospecting rather than retargeting, new customer sales rose 58%. That's the number that showed the extra reach was finding new buyers, not just reminding old ones.

Keep brand searches out of the result you scale on

If your Performance Max or Shopping results include people who searched your store's name, part of that ROAS was never at risk. Exclude your brand where you can, or at least report brand and non-brand separately. We've written about how brand searches flatter Performance Max in more detail.

Does your account have enough conversion data to scale?

Bidding strategies need data to work, and raising the budget doesn't give them any more of it on day one. Google sets minimums for Target ROAS bidding: Search and Shopping campaigns need at least 15 conversions in the past 30 days, and Demand Gen campaigns need at least 50 in the past 35 days.

• Below those levels, fix tracking and grow volume before you set an aggressive ROAS target. A target set on thin data mostly guesses.
• Raise budgets in steps, not leaps, and judge each step over a full buying cycle rather than a weekend.
• Remember daily spend can run hot. Google's budget rules let a campaign spend up to 2 times its average daily budget on a given day, while capping the month at 30.4 times the daily amount. A new budget can look alarming for a day or two without being wrong.

How to tell whether extra ad spend is adding sales

The real question about scaling isn't whether ROAS holds. It's whether the extra dollars bring extra orders that wouldn't have come anyway. Platforms call this incrementality, and it's measured by comparing people who saw your ads with people who didn't.

Google's Conversion Lift splits people into a treatment group that sees your ads and a control group that doesn't, and reports the difference as the lift. It isn't available on every account, so you'll often need a simpler version:

1. Watch your total store revenue against your total ad spend, week by week, not just the platform's reported revenue.
2. When you raise a budget, hold everything else still for long enough to see whether total orders move with it.
3. If platform-reported sales climb but store sales don't, the platform is claiming orders that would have arrived anyway.

This is also where an honest result can look less flattering. On On The Nose Co's account, paid search grew Google Ads purchase value 111% and return on ad spend 6% against the initial full-quarter baseline, although acquisition cost rose as activity expanded. That's normal when you scale, and it's exactly why you need to know your break-even before you start.

How to scale ecommerce ad spend once the numbers hold up

Once you know your break-even ROAS, your tracking matches your store, and you can see new customers separately, scaling becomes a series of small, readable tests rather than a leap of faith.

1. Grow the campaigns that find new customers, not the ones living off your brand name and repeat buyers.
2. Raise budgets in steps and judge each step against break-even and total store sales.
3. Add reach where the data supports it, such as non-brand search, new markets or prospecting audiences.
4. Stop at the point where the extra sales cost more than your margin allows, even if the platform's ROAS still looks fine.

For CheekyGlo's ad scaling, we built out non-branded campaigns and expanded Performance Max internationally. Campaign spend rose 110% while conversions rose 126% and overall ROAS rose 22%, which is what scaling on sound numbers looks like: the efficiency held as the budget grew. On FilamentHub's Google Ads, we assessed purchase value and acquisition costs as spend increased. Against the initial full-quarter baseline, Google Ads purchase value rose 158% and cost per purchase fell 35%.

If you sell through a product feed, the feed is part of this too. Google Shopping management starts with feed quality, because Shopping can only scale on the products Google approves and matches well. And if you want to see how other online retailers have handled it, our ecommerce results show the figures we've published for each store.

What to measure before scaling ad spend: the checklist in one table

None of this needs expensive software. It needs a margin you trust, tracking you've checked against your own orders, and the patience to raise budgets one readable step at a time. If you'd like a second pair of eyes on it, a growth marketing agency for ecommerce brands should start by checking exactly these numbers before it touches your budget. For the organic side of the same store, see our approach to SEO for fashion and ecommerce brands.