Introduction: Why GA4 Matters When You’re Running Amazon Ads in South Africa
A common South African scenario looks like this: you’re spending on Sponsored Products, maybe some Sponsored Brands, and the Amazon Ads console says orders are coming in. Yet when you check Google Analytics 4, traffic from Amazon appears fragmented, revenue is missing, and your broader media report cannot explain how Amazon influenced demand on the way to purchase. That gap matters because many SA sellers are not only trying to win on Amazon itself; they are also balancing Meta, Google Search, email, and marketplace traffic, all in ZAR, with VAT and logistics constraints that affect the real margin picture.
GA4 does not replace Amazon Ads reporting. Instead, it gives you a cleaner view of what happened before and after the click, especially when your Amazon activity is part of a wider acquisition system. For Prebo Digital, the practical value is not simply “more data.” It is using one measurement layer to compare paid channels, understand assisted conversions, and separate platform-reported success from business-level performance. That is especially useful for South African sellers who need to decide whether to scale Amazon Sponsored Products, tighten spend, or shift budget toward a channel that creates more profitable demand.
Info: GA4 is most useful here when it is configured as a reporting and attribution layer, not as a replacement for Amazon’s own advertising console.
Prerequisites: What Must Be in Place Before You Connect Anything
Before you touch tags or dashboards, confirm four basics. First, you need access to Amazon Attribution if you want Amazon traffic to carry trackable links into GA4. Second, GA4 must already be installed on the site or landing page where Amazon traffic lands. Third, Google Tag Manager should be available if you want a flexible setup that can scale without developer intervention every time a campaign changes. Fourth, you need to know which pages are actually measurable: Amazon product detail pages, Stores pages, and off-Amazon landing pages do not behave the same way, so the tracking plan has to match the destination.
The South African eligibility question matters too. Some Amazon ad features roll out differently by marketplace and seller setup, and your ability to use Amazon Attribution, Sponsored Products, Sponsored Brands, or Stores depends on account and marketplace access rather than your media budget alone. In practice, a local seller should verify whether the campaign is promoting a product detail page, a Store, or an external landing page, because the measurement method changes with each destination.
| Requirement | Why it matters | Practical note |
|---|---|---|
| Amazon Attribution access | Creates trackable links for off-Amazon measurement | Check marketplace eligibility before launch |
| GA4 property | Captures sessions, engagement, and conversions | Use consistent naming for channels and campaigns |
| Google Tag Manager | Simplifies tag deployment and updates | Best for teams that iterate often |
Step-by-Step Integration: UTM Tagging and Google Tag Manager
The cleanest setup starts with a naming system. Use Amazon Attribution links where possible, then append UTMs that make sense in GA4. The goal is not to create five versions of the same campaign name; it is to make source, medium, and campaign easy to interpret in one report. For South African teams, that also means keeping currency and market context visible in your reporting layer, especially if the rest of the business is already reviewed in ZAR.
A simple structure might look like this: source = amazon, medium = sponsored-products, campaign = winter-launch-sa. If you run off-Amazon traffic to a product landing page, make the destination page the same one you use in GA4 conversion analysis. That way, you can compare landing-page behaviour, not just clicks. In Google Tag Manager, make sure the GA4 configuration tag fires on every relevant page and that your key events are mapped before campaigns go live.
Example UTM structure for Amazon-supported traffic:
https://www.example.co.za/product-page?utm_source=amazon&utm_medium=sponsored-products&utm_campaign=sa-winter-launch&utm_content=headline-variant-a
Recommended naming rules:
- utm_source: amazon or amazon-dsp
- utm_medium: sponsored-products, sponsored-brands, sponsored-display, email, paid-social
- utm_campaign: business-readable launch name
- utm_content: creative or placement variantTip: Keep one campaign naming dictionary shared between media, analytics, and finance. That avoids mismatches when reporting in ZAR across Amazon and GA4.
Where Google Tag Manager fits in
GTM is the place to standardise event capture. If you track product views, add-to-cart actions, leads, or purchases on a brand site linked from Amazon campaigns, GTM helps you deploy those events without rebuilding code every time. This is useful when a South African team needs to test a new landing page despite limited development bandwidth or occasional load-shedding disruptions that slow live-site changes. In that environment, a tag-managed workflow is often more stable than ad hoc manual edits.
Reconciling Metrics: Why Amazon and GA4 Will Never Match Perfectly
Amazon Ads and GA4 use different measurement logic. Amazon Ads is built around ad exposure, click paths within Amazon, and attribution windows defined by the platform. GA4 is session-based and event-based, with its own attribution model and cross-channel view. If a shopper clicks an Amazon ad, browses, leaves, returns later via direct traffic, and converts, both systems may assign credit differently. That is normal and not a sign that one platform is broken.
| Metric view | What it answers | Typical limitation |
|---|---|---|
| Amazon Ads console | Which campaigns drove attributed sales on Amazon | Platform-specific attribution window |
| GA4 | How users behaved across sessions and channels | May undercount marketplace-native conversions |
| Combined reporting | How Amazon influenced total demand | Needs consistent naming and a shared definition of success |
A practical reconciliation rule is to compare trends, not raw totals. If Amazon clicks are up but GA4 engaged sessions are flat, your landing-page relevance, device mix, or traffic quality may need attention. If Amazon sales are up while GA4-assisted conversions are down, the issue may be attribution blindness rather than poor performance. The real task is to understand what each system is designed to measure.
Local Considerations: VAT, Load-Shedding, and Agency Support
South African reporting needs extra care. VAT treatment should be separated from media performance so your dashboard does not blur gross sales, net sales, and tax-inclusive values. If your finance team reports in ZAR, keep Amazon and GA4 exports aligned to the same currency view wherever possible. For many teams, that means building a monthly bridge between platform metrics and finance numbers rather than trusting a single dashboard to do both jobs.
Load-shedding also creates a practical issue: if staff are switching between platforms during unstable connectivity, incomplete tagging checks and late campaign updates become more common. That is why Prebo Digital typically recommends a documented launch checklist, role-based access, and a fallback approval process for tracking changes. For sellers working with an agency, the support model should include setup, QA, and reporting governance, not just ad buying. A strong partner will help you explain data discrepancies, keep tags tidy, and make sure campaigns are reviewed in business terms, not only in platform terminology.
