Across the e-commerce brands we manage, our team has put more than $123,000 into paid social and pulled back over $600,000 in tracked revenue. That is roughly a 5x blended return, with our best-performing account sitting north of 19x lifetime.

We are not telling you that to brag (okay, a little). We are telling you because that number, the ROAS number everyone chases, is the single most misread metric in paid advertising. And once you learn to read it the way we do, you stop lighting money on fire.

Here is what six figures of spend actually taught us.

Blended ROAS is the number that matters. Campaign ROAS is the one that fools you.

Open any ad account and you will find a campaign posting a beautiful 8x and another one barely scraping 1.5x. The instinct is to kill the 1.5x and pour everything into the 8x.

Sometimes that is right. Often it is a trap. That “underperforming” campaign might be the top-of-funnel that introduced the customer who later converted on the 8x retargeting campaign. Kill the intro, and the closer starves.

We manage to blended ROAS, the return across the whole account, because that is the number that actually hits the bank. A single campaign’s ROAS is a clue, not a verdict.

The platform is lying to you. A little. In both directions.

We run sites on WordPress with more payment processors than we can count, and we have watched Meta over-count conversions on one and quietly under-count on the next. TikTok does the same. The pixel is doing its best, but it is estimating.

This is why we never trust a single source. We reconcile the ad platform against Google Analytics and the store’s own order data before we make a call. Google, for what it is worth, tends to be the honest one: consistent, and you can usually trust the number it hands you. When your ad account and your actual sales disagree, your actual sales win. Every time.

Give the pixel time to warm up.

New accounts and fresh pixels get judged on a single day and killed too early constantly. The pixel needs conversion data before the algorithm knows who to find. Judging a campaign in its first 24 hours is like grading a book by its first sentence.

We build in a warm-up window before we read performance seriously. The clients who let the system learn are the ones sitting at 4x and 5x months later.

Scaling too fast is how you lose the account, not just the ROAS.

This one is counterintuitive: your ads can perform too well. When you go from a $120 cost-per-purchase to $7 and try to 10x the budget overnight, the platform gets suspicious. Sudden, aggressive scaling on top of a huge ROAS is one of the fastest ways to get an account flagged.

The move is to scale in deliberate steps and shift budget gently toward the winners, not to yank it all at once. Patience protects the asset.

The ad was never the whole story.

Here is the part that ties the $600k together. The ads did not generate that return on their own. The ad is the spark. The return comes from everything the spark lands on.

Paid traffic hitting a page built to convert instead of a homepage and a prayer. The visitor who did not buy getting captured into email and retargeting, so you are not paying twice for the same person. Owned channels turning one purchase into three. That is how ad dollars compound instead of leaking.

A great ROAS is not the finish line. It is what happens when the whole system is built to catch what the ads bring in.

The takeaway

If you are reading your ad account one campaign at a time, trusting the platform’s number at face value, and judging performance on day one, you are almost certainly leaving return on the table. Read blended. Reconcile your sources. Let it warm up. Scale with intention.

And if you would rather hand the whole thing to a team that has already made these mistakes so you do not have to, that is what we do. If you are curious what your accounts could be doing, we are happy to take a look.