A store owner checks last week’s sales: Shopify says 120 orders, Meta Ads Manager says its ads drove 70, Google Ads claims 55, and GA4 gives paid social 30 and paid search 40. Added up, the ads “caused” more sales than the business made.
Nobody is lying. Each tool answers a different question with different rules. Once you know the rules, you can use each number for what it’s good at.
The counting rules, side by side
| Meta Ads Manager | Google Ads | GA4 | |
|---|---|---|---|
| Credits | Meta ads only | Google ads only | All channels (paid, organic, email, referral…) |
| Default window | Standard attribution: 7-day click and 1-day view, plus 1-day engage-through (for example video plays) | Click-through: 30 days by default (1–90 configurable). View-through: 1 day. Engaged-view: 3 days | Lookback: 90 days for most key events (30 days for first visits) |
| Counts views? | Yes, 1-day view-through by default | Some view-through and engaged-view conversions, depending on campaign type | Mostly clicks and visits. Engaged views of YouTube ads can count. |
| Model | Last touch within Meta’s windows, or an “incremental” model if you choose it | Data-driven or last click, across Google ads | Data-driven (default), or paid and organic last click |
| Date of the conversion | Reported by ad impression date or by conversion date, depending on the report setting | “Conversions” reported by click date; “by conv. time” columns by conversion date | Date of the key event |
Sources: the platforms’ help pages listed below. Meta’s settings are chosen per ad set, so two ad sets can count differently.
Why the totals add up to more than reality
One customer can see an Instagram ad on Monday, search your name on Google on Wednesday, click a Google ad, and buy. Meta counts it (click or view within its window). Google Ads counts it (click within 30 days). GA4 gives the credit to Google paid search, or splits it, but counts the sale once. Your store counts it once.
Each ad platform only sees its own ads, so each takes full credit for shared customers.
Other reasons the numbers differ
- Time zones. Google’s help notes reports can differ if the Google Ads account and GA4 property use different time zones. Check every tool is set to UAE time (GMT+4).
- Reporting date. A sale on the 2nd from a click on the 30th lands in different months in different tools.
- Tracking gaps. Blocked browser tracking, a tag that fails, consent choices. Server-side tracking helps. See Conversions API and enhanced conversions.
- Modelled conversions. Platforms estimate some conversions they can’t observe directly. Totals can change after the fact.
- Duplicates. A purchase event that fires twice (on refresh of the thank-you page, or Pixel and server without deduplication).
- Refunds and cancellations. Your store subtracts them. Ad platforms usually don’t, unless you send them.
Which number to use for what
| Question | Use |
|---|---|
| How many sales or leads did we actually get? | Your store or CRM. This is the truth for totals. |
| Which campaign or ad works better within Meta? | Meta Ads Manager, comparing like with like (same attribution setting) |
| Which keyword or campaign works better within Google Ads? | Google Ads |
| How do channels compare with each other? | GA4, with the caveat that it under-credits views and social |
| Are ads making sales that wouldn’t have happened anyway? | A holdout or incrementality test (below) |
A useful habit: track a blended cost per sale every week, meaning total ad spend divided by total new sales from your own system. If it rises while platform numbers look great, something is being over-counted.
Testing real impact: a simple holdout test
Attribution tells you who gets credit. It doesn’t tell you whether the ads caused the sale. For that you need a comparison:
Platform lift tests
Meta and Google both offer lift or conversion-lift studies in some accounts, which hold back a random group from seeing ads. Eligibility depends on spend and account, so check what your accounts offer.
A do-it-yourself geographic holdout
For a business selling across several areas:
- Pick two comparable areas (for example two emirates, or groups of cities) with stable sales.
- Record 4 weeks of sales in both, with ads running in both.
- Pause ads in one area (the holdout) for 2–4 weeks, and keep everything else the same.
- Compare: did sales in the holdout fall relative to the area that kept ads?
The difference, compared with the spend you saved, is a rough measure of what the ads really add. It’s imperfect (areas differ, seasons change) but far better than trusting attributed numbers alone. Avoid holidays and big promotions during the test.
Meta also offers an incremental attribution model, which predicts whether a conversion was caused by an ad and optimises for that. Meta’s note: results under different attribution models can’t be compared directly.
Frequently asked questions
Should we change Meta’s attribution setting to match Google?
You can choose a different setting per ad set, but there’s no setting that makes them agree. Pick one setting, keep it consistent so trends are comparable, and judge totals on your own data.
Why did GA4 show fewer conversions than Google Ads for the same campaign?
GA4 shares credit across channels and Google Ads only credits its own ads. Differences in windows, time zones, and the columns you’re comparing (“Conversions” vs “All conv.”) also matter.
Do we need expensive attribution software?
Rarely, at small-business scale. Clean tracking, consistent UTMs (see UTM naming conventions), your own sales data as the total, and an occasional holdout test answer most questions.
If you’d like your reporting set up so the numbers are explainable, that’s part of our traffic sources service.
Free download · Excel workbook
Attribution reconciliation sheet
Record weekly conversions from Meta, Google Ads, GA4 and your own orders or CRM side by side, see the gaps, and plan a simple holdout test with a before-and-after comparison.
Download itNo sign-up. Opens in Excel, Google Sheets or Numbers.
Rather have it done?
This is the kind of work our traffic sources service covers. Tell us where you are today and we’ll reply with a plan and a fixed price.
Let’s talk