How to Track Ad Spend and ROAS Across Meta and Google

You spent $4,200 on Meta last month. $2,800 on Google. Revenue went up. But which platform drove the revenue? And how much of that revenue would have happened anyway?

Most founders look at total spend, total revenue, and divide. That gives you a blended number that hides everything useful. A 3x blended ROAS sounds great until you realize Meta is running at 1.4x and Google is carrying the whole average at 5.8x. You're lighting money on fire on one platform and you can't see it because the other one is compensating.

Tracking ad spend properly is not complicated. But it requires you to stop looking at aggregate numbers and start looking at each channel on its own terms.

The ROAS Formula (and Why People Get It Wrong)

ROAS = Revenue attributed to ads / Ad spend. That's it. If you spent $1,000 and those ads generated $4,000 in revenue, your ROAS is 4.0x.

Where founders trip up: they use total revenue instead of attributed revenue. If your store made $40,000 last month and you spent $5,000 on ads, your ROAS is not 8x. Some of that revenue came from organic traffic, email, direct visits, repeat customers who would have bought anyway. You need to isolate what the ads actually brought in.

Both Meta and Google have attribution models built into their dashboards. They're not perfect. Meta tends to over-attribute because it counts view-through conversions by default. Someone scrolled past your ad, bought your product three days later through a Google search, and Meta takes credit. Google is more conservative on view-throughs but still has its biases.

The pragmatic approach: use each platform's own attribution for per-channel ROAS, but calculate blended ROAS separately using your actual total revenue minus organic baseline. We'll get to blended ROAS in a minute.

What's a Good ROAS? Real Benchmarks by Platform

The honest answer is that “good” depends on your margins. A 3x ROAS on a product with 80% gross margin is incredible. A 3x ROAS on a product with 30% gross margin means you're losing money after fulfillment and overhead.

That said, here are the numbers that actually reflect what businesses see across platforms in 2026:

Meta Ads (Facebook + Instagram): Average ROAS across industries sits around 2.2x. Ecommerce brands with strong creative can push 3x to 4x. Below 1.8x, you're likely underwater unless you have very high margins or a strong LTV play where you're willing to lose money on the first purchase. The creative matters more than the targeting now. Meta's algorithm is excellent at finding buyers if you feed it good creative. Bad ad, bad ROAS. It's that direct.

Google Ads (Search + Shopping): Average ROAS is around 4.5x, but this is heavily skewed by branded search. If someone searches your brand name and clicks your ad, that's a near- guaranteed conversion. Strip out branded terms and your non-brand search ROAS is probably closer to 2.5x to 3.5x. Shopping campaigns for ecommerce tend to run between 3x and 6x depending on the category and competition. Performance Max is a black box that Google claims outperforms everything. Sometimes it does. Sometimes it's spending your money on Display Network placements nobody sees.

YouTube Ads: Generally 1.5x to 2.5x for direct response. Better for awareness than for immediate ROAS, which makes it hard to justify if you're measuring strictly on last-click attribution.

Blended ROAS: The Number That Actually Matters

Per-channel ROAS tells you which channels are efficient. Blended ROAS tells you whether your overall marketing machine is working.

Blended ROAS = Total revenue / Total ad spend across all channels. This gives you the big picture. If you're spending $10,000 across Meta and Google combined and generating $35,000 in total revenue, your blended ROAS is 3.5x.

The reason blended ROAS matters: channels work together. Someone sees your Meta ad, doesn't click. Googles your brand two days later, clicks your Google ad, buys. Meta gets zero credit in platform reporting. Google gets full credit. In reality, both channels contributed. Blended ROAS captures this interplay without trying to solve attribution perfectly.

A healthy blended ROAS for most ecommerce businesses is 3x or above. For SaaS with high LTV, even 1.5x can work if your payback period math holds up. The target depends entirely on your unit economics.

How to Compare Channels Without Fooling Yourself

Comparing Meta and Google head-to-head on ROAS alone is misleading. Google Search captures intent that already exists. Someone typed “buy running shoes size 10.” They were going to buy from someone. Meta creates demand that didn't exist. Someone was watching cooking videos and your ad interrupted them with shoes they didn't know they wanted.

These are fundamentally different jobs. Judging them by the same ROAS threshold is like comparing a closer to a lead generator. You need both.

What works better: set a minimum ROAS threshold per channel that accounts for its role. Maybe Google Search needs to hit 4x because it's pure intent capture. Maybe Meta only needs to hit 2x because it's filling the top of your funnel. As long as your blended ROAS stays healthy, the individual channel thresholds should reflect the job each channel does.

The Tracking Setup That Actually Works

UTM parameters on every ad link. Every single one. If you're running ads without UTMs, your analytics tool has no idea where the traffic came from. Use a consistent naming convention. Something like utm_source=meta&utm_medium=paid&utm_campaign=spring-sale that your whole team follows.

Set up conversion tracking properly on both platforms. Meta's Conversions API (server-side) in addition to the pixel. Google's enhanced conversions. iOS privacy changes killed a lot of pixel-only tracking, so server-side is no longer optional.

Then pull your spend and attributed revenue into one place. This is where most founders lose the thread. The data lives in three dashboards: Meta Ads Manager, Google Ads, and your analytics tool. You end up switching between tabs, copying numbers into a spreadsheet, and losing an afternoon every week.

Kartib pulls in your Meta Ads and Google Ads data into a single dashboard. You see per-channel ROAS, blended ROAS, spend trends, and cost-per-acquisition without jumping between platforms. It also flags when a channel's ROAS drops below your threshold so you catch problems before the budget is gone.

When to Kill a Channel

If a channel is consistently below your minimum ROAS threshold for four consecutive weeks, cut spend by 50% and test new creative. If it still doesn't recover after two more weeks, pause it entirely and reallocate to what's working.

The exception: new channels need time to learn. Meta's algorithm needs about 50 conversions per week to optimize properly. If you're only getting 5 conversions a week, the ROAS data isn't reliable yet. You're not measuring the channel. You're measuring noise.

Don't be sentimental about channels. Facebook was the best acquisition channel for DTC brands for years. For some brands, it still is. For others, it stopped working in 2024 and they're still spending because they remember when it was good. Your ROAS data tells you what's true right now.

The Bottom Line

Track per-channel ROAS to know which platforms are efficient. Track blended ROAS to know whether your overall acquisition machine is profitable. Set different thresholds for different channels based on their role in the funnel. And get your spend data into one place so you can actually see what's happening without wasting time in three different dashboards.

The founders who scale paid acquisition successfully are not the ones spending the most. They're the ones who see their numbers clearly, react fast when something breaks, and never let a bad channel run on autopilot.