GA4 Data Manager: what changes when you blend Meta, TikTok and GA4

Data Manager in GA4 is Google's move, announced September 10, 2026, to plug its first-party data tool (CRM, offline sales, app data) straight into Google Analytics. The same announcement brought enhanced conversions to GA. If you're blending Meta and TikTok with GA4, though, the piece that brings those platforms' costs in is a different one: campaign data import, with automated connections for Meta, TikTok, Snap, Pinterest and Reddit. Together they push GA4 closer to an all-channel dashboard. The limit hasn't moved: GA4 only sees a CRM sale if you send it there.

What Google announced

On September 10, 2026, Google published a data and measurement update. The parts that matter for GA4:

• Data Manager inside Google Analytics (and Display and Video 360), to manage and activate first-party data across tools.
• Enhanced conversions in GA and DV360, which use customer-provided data, hashed, to match conversions to people who saw your ads.
• Data Strength Uplift, a metric estimating how many conversions your first-party data setup recovered.

Google also shared numbers, such as an average 26% lift in incremental ROAS for advertisers who connect offline and app data through Data Manager. That's Google's own figure, from Google's own data, on Search campaigns. Treat it as direction, not a forecast for your account.

Three pieces, kept apart

Most of the confusion comes from calling everything "Data Manager". There are three things:

Data Manager isn't where you connect Meta Ads. Meta's cost comes in through campaign data import, in GA4 Admin. Mixing the two up is how teams end up setting up a Data Manager connection and then wondering why Meta spend never shows in their reports.

How Meta and TikTok costs get into GA4

Campaign data import (formerly "cost data import") pulls cost and impressions from linked platforms and joins them to GA4 sessions. The join uses three keys: utm_source, utm_medium and date. That's where the rules that break most imports come from:

1. UTMs must match exactly. The source and medium you enter in GA4 must be identical to the ones in the ad URL. "facebook" and "Facebook" are different values. The naming scheme that prevents this is in UTM parameters for ads.
2. Dynamic parameters don't work. GA4's import setup doesn't accept dynamic placeholders. Resolve them at the ad platform.
3. Don't mix the automated connection with manual files for the same platform. GA4 doesn't de-duplicate between the two.
4. Connections can't be edited. Account, source and medium are locked once created. Got one wrong? Delete it and start over.

Since August 10, 2026, there's a campaign data import validation report that flags campaigns missing cost, clicks or impressions. Run it before you trust any channel-level ROAS in GA4.

What changes if you blend Meta, TikTok and GA4

Until now, the blending happened elsewhere: each platform's cost in one sheet, GA4 sessions and conversions in another. With import and Data Manager, GA4 can show cost, conversions and ROAS by channel on one screen, under one attribution model for all of them.

That settles one argument and starts another. It settles double counting: GA4 won't credit the same sale to both Meta and Google, the way each platform does on its own side. It starts the mismatch: Meta's ROAS in GA4 won't match Ads Manager, because GA4 attributes by session under its own model, and Meta attributes by click and view inside its own window. Why the windows differ is covered in attribution windows explained.

The working rule: use GA4 to compare channels against each other, under one yardstick. Use each platform to make decisions inside it, because that's the number its bidding uses.

The limit: GA4 still stops at the lead

If you sell over the phone, by chat or through a sales team, the sale happens in the CRM, days or weeks after the click. GA4 doesn't see that on its own. It sees the form submission.

Data Manager is exactly the route for getting that sale to Google, but it takes work:

• A key that links the sale to the click or the person. Google's click identifier, or the email and phone the customer gave you, sent hashed.
• CRM discipline. Sales marked at the right stage, on the right date, with a value. A messy CRM turns into wrong conversions in GA4, and the bidding learns from them.
• Patience with the lag. A sale that closes in 20 days shows up 20 days later. Long cycles and how to measure them are covered in offline attribution.

And there's the part no Google integration covers: your CRM sale goes to Google, not to Meta or TikTok. For those platforms, you send it through each one's own conversions API.

Starter checklist

• Standardize utm_source and utm_medium across every Meta and TikTok ad.
• Create one automated campaign data import per platform, with no manual file running in parallel.
• Run the validation report after 48 hours.
• Turn on user-provided data collection if you'll use enhanced conversions.
• Decide the sale's matching key in your CRM before opening Data Manager.
• Agree with the team which number answers which question, so the argument doesn't restart every month.

Where Vazante fits

Next step

If your GA4 doesn't have Meta and TikTok costs yet, start with campaign data import: it's the biggest payoff for the least effort. Then decide your CRM matching key, and only then bring offline sales in through Data Manager. If your CRM isn't wired to your ads at all yet, start with connect Meta Ads to your CRM. To see all of it side by side in Claude, the route is paid ads reporting.