You check Meta Ads Manager: 50 conversions this week. You check Google Ads: 30 conversions. You pull up Shopify: 45 orders. None of the numbers match. They never match. And somehow, the total across all platforms is 80, which is nearly double what you actually sold.
This isn’t a bug. It’s how the system works. Here’s why, and what to do about it.
The Core Problem: Each Platform Tells Its Own Story
Meta, Google, TikTok, and every other ad platform are each watching the same customer journey through a different window. They each see some of the picture but never all of it. And they each have a strong incentive to take credit for as many conversions as possible.
This isn’t dishonest. It’s a natural consequence of how attribution works when multiple platforms are involved and none of them can see each other’s data.
To understand why the numbers disagree, you need to understand four things: attribution windows, counting methods, click definitions, and cross-device tracking.
Attribution Windows: When Does a Conversion “Count”?
Each platform has a different default window for how long after an interaction they’ll claim credit for a conversion.
| Platform | Click Window (Default) | View Window (Default) |
|---|---|---|
| Meta Ads | 7 days | 1 day |
| Google Ads | 30 days | 1 day |
| TikTok Ads | 7 days | 1 day |
| Pinterest Ads | 30 days | 1 day |
Here’s what this means in practice.
A customer clicks a Meta ad on March 1st. They don’t buy anything. On March 5th, they Google your brand name, click a Google ad, and make a purchase.
Meta’s perspective: The customer clicked our ad 4 days ago and then converted. That’s within our 7-day click window. We claim this conversion.
Google’s perspective: The customer clicked our ad today and converted. Obviously that’s our conversion.
Both platforms are telling the truth according to their own rules. Both claim the same single purchase. Your Shopify dashboard shows 1 order. Your ad platforms, combined, show 2 conversions.
Now imagine this happens dozens of times a week across thousands of customers. That’s why the numbers never add up.
View-Through Conversions: The Silent Multiplier
View-through attribution is where things get especially messy. Meta counts a conversion if someone merely saw your ad (in their feed, didn’t click it) and then converted within 1 day. They don’t have to click. They just have to have been shown the ad.
Meta’s 1-day view default is significant. Meta shows billions of ads per day. If someone scrolls past your ad in their feed at 8am and then buys from your site at 6pm, Meta claims that conversion. The customer may not even remember seeing the ad.
Google Ads does offer view-through for Display and YouTube campaigns, but it’s less aggressive by default and not applied to Search campaigns at all.
This is one of the biggest reasons Meta often reports more conversions than Google for the same campaign period. Meta is counting events that Google would not.
To check your view-through impact in Meta: Go to Ads Manager, customize your columns, and separate “1-day view” conversions from “7-day click” conversions. You might be surprised how many of your reported conversions are view-through.
Counting Methods: What Counts as “One Conversion”?
Beyond attribution windows, the platforms disagree on what a “conversion” even means.
Meta: Counts Every Conversion
By default, Meta counts every purchase event. If one person makes 3 separate purchases after clicking your ad, Meta reports 3 conversions.
Google Ads: Configurable Per Action
Google Ads lets you choose between “Every” and “One” for each conversion action. The default depends on the type of conversion. Purchases usually default to “Every” (like Meta), but lead form submissions might default to “One” per user.
GA4: Session-Based by Default
GA4, which many businesses use as a cross-reference, has its own quirks. It historically counted one conversion per session per event, meaning 3 purchases in a single session would only count as 1 conversion. (Google has been updating this, but the behavior depends on your configuration.)
These different counting methods mean that even if both platforms saw the exact same events, they might report different totals.
Click Definitions: Not All Clicks Are the Same
When Meta says someone “clicked” your ad, what does that mean?
Meta counts several types of interactions as clicks, including link clicks, but also engagement clicks like comments, shares, and reactions on certain ad formats. The metric you want to compare is “link clicks” (also called “outbound clicks”), not “clicks (all).”
Google Ads counts clicks as clicks to your website. It’s more straightforward, but Google also deduplicates. If someone accidentally double-clicks your ad, Google typically counts it once.
These differences are small individually, but they contribute to the overall discrepancy in conversion reporting because each platform’s click pool (the set of interactions that start an attribution window) is defined differently.
Cross-Device Tracking: The Identity Problem
Here’s where it gets really interesting. People don’t use one device. They see an ad on their phone during lunch, research on their laptop after work, and buy on their tablet from the couch.
Each platform tries to stitch these sessions together, but they can only do it for users they can identify.
Google’s advantage: Google has logged-in users across Chrome, Gmail, YouTube, Android, and Google Search. If you’re signed into Chrome on your phone and your laptop, Google can connect those sessions.
Meta’s advantage: Meta has logged-in users across Facebook, Instagram, WhatsApp, and Messenger. If you’re logged into Instagram on your phone and Facebook on your laptop, Meta can connect those.
The gap: Neither platform can see the other’s data. Google can’t see that a user interacted with your Meta ad on Instagram. Meta can’t see that the same user searched your brand name on Google. Each platform only has its piece of the puzzle.
This means cross-device conversions get attributed differently by each platform, and some conversions fall through the cracks for both.
The Math: Why Totals Always Exceed Reality
Let’s work through a concrete example.
You run a Shopify store. This week you had 45 actual orders.
Your ad spend:
- Meta Ads: $3,000/week
- Google Ads: $2,000/week
Your platforms report:
- Meta Ads: 38 conversions
- Google Ads: 27 conversions
- Total claimed: 65 conversions
But you only had 45 orders.
What happened? Some of those orders were touched by both platforms. A customer saw a Meta ad, then Googled your brand name and clicked a Google ad, then bought. Both Meta and Google claim that conversion. Neither platform is lying. They’re each correctly reporting based on their own attribution rules.
The 20-conversion “overcounting” (65 claimed minus 45 actual) represents the overlap. Those are real customers who interacted with both platforms before purchasing.
This is normal. It happens to every business running ads on multiple platforms. The overlap is typically 20-50% of total claimed conversions.
Which Number Should You Trust?
The short answer: none of them individually. Each number tells you something different.
For Campaign Optimization Within a Platform: Trust That Platform
Meta’s algorithm optimizes using Meta’s conversion data. If Meta reports 38 conversions, that’s what its algorithm sees, and that’s what it uses to find more customers like those 38. Changing Meta’s conversion data source mid-campaign can disrupt optimization.
The same applies to Google. Use Google’s data for Google campaign decisions.
For Budget Allocation Across Platforms: Trust Your Backend
This is where people make the most expensive mistakes. If Meta reports 38 conversions and Google reports 27, it’s tempting to think Meta is performing better and shift budget there. But if you dig into your backend data:
- Orders from customers who only interacted with Meta ads: 20
- Orders from customers who only interacted with Google ads: 10
- Orders from customers who interacted with both: 15
The picture is more nuanced. Cutting Google would lose those 10 single-touch orders and potentially some of the 15 multi-touch orders that Google assisted.
For Actual Business Performance: Trust Your CRM or Shopify
Neither Meta nor Google knows about:
- Cancelled orders
- Returns and refunds
- Fraudulent transactions
- Repeat purchases from existing customers
- Orders placed by phone or in person
Your backend (Shopify, WooCommerce, CRM, database) is the only source that reflects reality.
How to Reconcile the Numbers
You can’t make the numbers match perfectly, but you can build a framework for understanding the gap.
Step 1: Establish Your Baseline
Pull your actual backend data: total orders, total revenue. This is your ground truth.
Step 2: Calculate Each Platform’s Over-Count Ratio
Divide each platform’s reported conversions by your actual total:
Meta over-count ratio = Meta reported / Actual orders
Google over-count ratio = Google reported / Actual orders
Example:
Meta: 38 / 45 = 0.84 (Meta claims 84% of your orders)
Google: 27 / 45 = 0.60 (Google claims 60% of your orders)
Combined: 65 / 45 = 1.44 (44% overlap/overcount)Step 3: Track These Ratios Over Time
If Meta consistently claims 80-85% of your orders, that ratio becomes a useful conversion factor. If the ratio suddenly jumps to 120%, something changed (likely a tracking issue or attribution setting change).
Step 4: Use UTM Parameters as a Cross-Reference
UTM parameters in your ad URLs let your analytics platform (GA4, Shopify analytics) independently track which platform drove each session. Use our UTM builder tool to create consistently tagged URLs. This gives you a third-party view of the data that doesn’t rely on either ad platform’s attribution.
It’s not perfect (last-click only, doesn’t capture view-through), but it’s a useful gut check.
How Server-Side Tracking Helps
Server-side tracking doesn’t eliminate attribution discrepancies, but it reduces them in two important ways.
Consistent Event Delivery
When your server sends conversion events to all platforms, it sends the same event data to each one at the same time. This ensures that both Meta and Google are at least working from the same set of events, even if they attribute them differently.
Without server-side tracking, ad blockers might prevent Meta’s pixel from firing while Google’s tag works fine. That creates an artificial discrepancy on top of the natural attribution differences.
Deduplication IDs
Server-side implementations typically include a unique transaction ID (like your order number) with every event sent to every platform. This helps with two things:
- It prevents the same conversion from being counted twice by a single platform (if both your browser pixel and server event fire)
- It gives you a clean key to reconcile across platforms and your backend
Better Match Rates
By sending hashed customer data (email, phone) directly from your server, platforms can match conversions to ad interactions more accurately. Higher match rates mean fewer “modeled” conversions and more precise reporting.
Practical Advice: What to Actually Do
Don’t Add Platform Numbers Together
If Meta says 38 and Google says 27, you didn’t get 65 conversions. You probably got somewhere around 45. Adding them together overstates performance and will lead to bad budget decisions.
Don’t Cut a Platform Because It Shows Fewer Conversions
A platform reporting fewer conversions might still be driving significant value. Google Search ads often show fewer conversions than Meta because they tend to be the last click before purchase, while Meta often gets credit for the first touch. But that last click is still essential to the conversion happening.
Watch the Trends, Not the Absolutes
The absolute numbers from any single platform are unreliable for cross-platform comparison. But the trends within each platform are useful. If Meta’s conversions drop 30% week over week while your actual sales are flat, something is wrong with your Meta tracking, not your ads.
Set Up a Simple Reconciliation Spreadsheet
Once a week, pull three numbers: Meta conversions, Google conversions, and actual orders. Track the ratios. Over time, you’ll develop an intuitive sense of what “normal” discrepancy looks like for your business. When something deviates, you’ll catch it fast.
Review Attribution Settings Periodically
Check your attribution windows and counting methods in each platform at least once a quarter. Platform defaults change, and a settings update can shift your reported numbers without anything actually changing about your business.
When the Discrepancy Signals a Real Problem
Some discrepancy is always expected. But certain patterns indicate a real issue:
- Platform reports more conversions than actual sales: Possible double-counting, incorrect conversion events, or test transactions being counted.
- Platform reports dramatically fewer conversions than expected: Tracking is broken. Check that your pixel/tag is firing correctly.
- Discrepancy changes suddenly: An attribution window changed, a pixel was accidentally removed, or consent mode kicked in without you realizing it.
- One platform’s numbers are wildly inconsistent week to week while sales are stable: The tracking for that platform is unreliable.
The Bottom Line
Meta and Google will never agree on your conversion numbers. That’s not a problem to solve. It’s a reality to manage.
Your job is to understand why they disagree (attribution windows, counting methods, view-through, cross-device), use each platform’s data appropriately (optimize within platform, allocate across platforms using backend data), and maintain a consistent source of truth (your CRM, your Shopify dashboard, your database).
Once you accept that no single platform has the complete picture, you can stop chasing matching numbers and start making better decisions with the data you have.
For the Google-specific version of this issue, see why GA4 and Google Ads show different conversions.
Get a free scan and we’ll check whether your conversion tracking is set up correctly across both Meta and Google, identify any deduplication issues, and show you exactly where your data gaps are.