Cross-Platform Attribution: Why Your ROAS Is Probably Wrong

The Attribution Problem That's Costing You Real Money

You're running Google Ads, Meta, and TikTok simultaneously. A user sees your TikTok ad on Monday. On Wednesday, they click a Meta retargeting ad. On Friday, they search your brand name, click a Google Search ad, and buy.

Who gets credit?

• Google Ads: Claims full last-click credit
• Meta: Claims an assisted conversion (7-day click window)
• TikTok: Claims a view-through conversion (1-day view window)

Your platform dashboards now show 2.5x your actual conversion count. Your reported blended ROAS looks incredible. You're reporting up to your CMO with confidence.

But it's fictitious. And the decisions you're making based on it are slowly degrading your actual performance.

Why Every Platform Over-Reports Conversions

This isn't an accident - it's structurally incentivized. Each platform's attribution system is optimized to make the platform look as essential as possible.

Google Ads defaults to data-driven attribution, which uses machine learning to assign credit across touchpoints. It sounds sophisticated, but it's still a Google-owned model that tends to favor Google-owned touchpoints.

Meta uses a 7-day click / 1-day view attribution window by default. This means if someone clicked a Meta ad within 7 days of converting, Meta claims credit - regardless of what else happened in between.

TikTok uses a 7-day click / 1-day view window, similar to Meta. For view-through attribution, a user just needs to have seen your ad in the past 24 hours.

When a user interacts with all three platforms (which is common), all three report a conversion. You might have 100 actual purchases but see 220 conversions reported across your platform dashboards.

The Real-World Impact on Your Budget Decisions

If you optimize based on platform-reported ROAS, you'll systematically make three expensive mistakes:

1. You'll over-invest in last-touch channels and under-invest in top-of-funnel

Google Search gets credit for conversions that Meta and TikTok seeded. You'll pour budget into Google, see diminishing returns, and wonder why Meta or TikTok "stopped working" after you cut them.

2. You'll pause campaigns that are actually driving value

A TikTok campaign with 0.6x reported ROAS might be driving significant branded search volume. Pause it, and your Google Search conversions drop a month later.

3. You'll attribute business growth to the wrong channels

This creates strategic errors. If you think 80% of revenue is coming from Google Ads, you'll hire more PPC specialists. If it's actually Meta doing heavy lifting, you're understaffed on social.

Four Ways to Fix Attribution

1. Use Marketing Efficiency Ratio (MER) as Your North Star

MER = Total Revenue / Total Ad Spend

This is the most robust metric available because it bypasses attribution entirely. You're measuring the aggregate efficiency of your entire ad investment. If MER is healthy, your ads are working. If MER is declining while platform-reported ROAS looks fine, something is wrong structurally.

Track MER week-over-week and month-over-month. It's the clearest signal you have.

2. Implement Conversions API (CAPI) on Meta and TikTok

Browser-based pixel tracking is increasingly unreliable due to iOS privacy restrictions, ad blockers, and cookie deprecation. Server-side Conversions API sends conversion events directly from your server to the platform - bypassing the browser entirely.

CAPI improves the signal quality Meta and TikTok receive, which improves their modeling. It doesn't solve the double-counting problem, but it improves data accuracy significantly.

3. Use Google Analytics 4 as a Neutral Source

GA4 isn't perfect, but it's not a platform with financial incentives to over-report. Its cross-channel attribution model gives you a more neutral view of the customer journey.

Compare GA4 conversions against platform-reported conversions for each channel. The gap tells you how much inflation each platform is claiming.

4. Run Incrementality Tests

This is the gold standard, but also the most effort. The method:

• Divide your audience into a test group (sees ads) and a holdout group (doesn't see ads)
• Measure conversion rate difference between the groups
• The difference = true incremental lift attributable to your ads

Most platforms (Meta, Google) have built-in lift study tools. They require meaningful spend levels to be statistically valid, but the insights are definitive.

The Attribution Stack That Works in 2026

You don't need all five layers to start making better decisions. MER tracking alone - which you can set up today - will immediately give you a more accurate picture than relying on platform dashboards.

Practical Next Steps

1. Today: Calculate your blended MER for the last 90 days. (Total revenue from your analytics / Total ad spend across all platforms)
2. This week: Compare that MER against each platform's reported ROAS. Spot the inflation.
3. This month: Connect all platform data to a single dashboard (MarketingSync makes this straightforward). Start tracking MER weekly.
4. This quarter: Run one incrementality test on your highest-spend channel to validate its true contribution.

The bottom line: Attribution in 2026 is a measurement problem, not a technology problem. The data exists. The challenge is aggregating it correctly and making decisions based on business outcomes rather than platform vanity metrics. MarketingSync pulls data from all platforms and lets you calculate true blended MER alongside platform-reported ROAS - so you always know the difference.