Your ROAS is lying to you (and the number that isn't)

I've sat across from founders who pull up the dashboard, point at a 4x return, and then quietly admit the bank balance hasn't moved in months.

Those two things should always agree. For a lot of brands, they don't. Once you understand why, it gets hard to fully trust ROAS again.

Every platform takes credit for the same sale

ROAS is built on revenue a platform claims it drove. The problem is that every platform claims the same purchase.

A customer sees your Meta ad, later clicks a Google branded-search ad, then converts after an email. Meta counts the sale. Google counts the sale. Klaviyo counts the sale. Add up the three dashboards and you've "driven" 3x the revenue you actually made.

Stack a few common defaults on top of that and the number drifts even further from reality:

• View-through conversions: someone saw the ad, didn't click, bought anyway, and the platform takes credit.
• Branded search: people who already knew you, searching your name, counted as paid acquisition.
• Direct and organic leaking into the attribution window.

We audited a new partner's account last year that reported a tidy 4.2x ROAS. After we deduplicated events and stripped out non-incremental revenue, the real number was 1.8x. Same spend, same product. The account had just been grading its own homework.

The number that agrees with your bank account

The fix isn't a better attribution model. It's looking at numbers that are hard to inflate:

• MER (Marketing Efficiency Ratio): total revenue ÷ total marketing spend, across every channel. One number, no double-counting, and it moves in step with your actual revenue.
• New-customer CAC: what it costs to acquire a new customer, not to re-buy one you already had. Branded search and retargeting can't hide here.
• Contribution margin after ad spend: the dollars left to run the business once you back out COGS and media.

These are less flattering than a platform-reported 4x. They're also the ones that match your bank account.

How to clean it up

1. Pick one source of truth for revenue (your store or a blended view), and judge channels against MER, not against each platform's self-report.
2. Fix the pixel. One clean purchase event, deduplicated. Three events firing on checkout is the single most common cause of inflated ROAS we see.
3. Tighten attribution windows. If your sales cycle is under 48 hours, test 1-day-click attribution. More conservative data is more trustworthy data.
4. Separate new from returning. Report new-customer ROAS alongside blended. If new-customer acquisition is unprofitable, you're renting growth, not building it.

If your ROAS looks great but the business doesn't feel like it's growing, your attribution is almost certainly broken. The good news: that's a fixable problem, and usually a fast one.

This is the kind of thing an expert who lives in your account 40 hours a week catches in week one, not month six. If you want someone who owns your attribution end-to-end, get matched with a Distilled Expert and tell us what your dashboards say. We will tell you what we would check first.

Related

• Hire a paid social expert
• How Mumzworld rebuilt Snapchat ROAS
• Somewhere vs Distilled