Is it true that implementing AMP inflates your direct traffic in Analytics?

How do AMP pages alter standard tracking?

The AMP system relies on a proxy cache hosted on Google's servers (cdn.ampproject.org). When a user clicks on an AMP result in mobile SERPs, the page doesn't load directly from your domain but from this cache.

The issue: this intermediate layer breaks the HTTP referrer, which normally allows Analytics to identify Google as the traffic source. As a result, the session appears as direct access, as if the user had typed your URL into their browser. Consequently, your acquisition reports become misleading, showing an abnormally high direct channel and an undervalued organic channel.

Why is this incorrect attribution a problem?

Direct traffic is traditionally associated with brand awareness, recurring users who directly type your URL, or access via favorites. If your AMP pages artificially inflate this channel, you lose the ability to properly measure these brand awareness metrics.

You also risk underestimating the actual impact of your SEO efforts. If 30% of your mobile organic traffic is classified as direct due to AMP, your management may wrongly conclude that organic search is performing poorly, when in fact it's just a tracking issue. The resulting budgetary decisions can be catastrophic.

Does this phenomenon affect all AMP sites?

No, the extent of the problem depends on your technical setup and your Analytics implementation. If you've correctly configured the data-linker attribute in your amp-analytics tag, the domain linker transmits the client ID between the hidden AMP page and your main domain, which can mitigate the issue.

But let's be honest: most of the sites I've audited have a shaky AMP implementation on this specific point. Development teams focus on AMP validation and performance, not on the Analytics plumbing. The result: faulty data for months until the problem is identified.

• Google's AMP cache breaks the natural transmission of the HTTP referrer
• Mobile organic traffic via AMP appears as direct traffic in Analytics
• This incorrect attribution distorts SEO performance and brand awareness reports
• The domain linker (data-linker) can mitigate but not fully resolve the issue
• Proper configuration requires close coordination between SEO and dev teams

Is Google’s explanation complete or watered down?

Google speaks the truth about the phenomenon, but seriously underplays its scope. Describing it as a "sudden increase" suggests it's a small temporary spike that you'll easily spot. In reality, if you implement AMP across all your mobile pages (as Google strongly encouraged a few years ago), it's not an isolated anomaly but a permanent distortion of your data.

I have seen e-commerce sites lose all ability to correctly measure their mobile acquisition funnel for over a year before understanding the source of the problem. The SEO ROI seemed to plummet while actual traffic increased. [To be verified]: Google provides no data on the average percentage of misattributed organic traffic based on implementations.

Do technical solutions really work?

The domain linker helps, but doesn’t solve everything. It transmits the client ID between domains, which allows you to link sessions, but doesn’t magically restore the referrer. You still have traffic classified as direct, although you can later trace it back to the source in deeper analyses.

The cleanest solution I've seen is to use custom UTM parameters on AMP URLs, combined with an advanced segment in Analytics to correctly reattribute this traffic. But it’s cumbersome to maintain and frankly, how many sites have the resources for that? The reality on the ground: most accept this loss of precision rather than invest weeks of development.

Should you still implement AMP considering this issue?

The real question. AMP is no longer a direct ranking factor since Google broadened the criteria for the Top Stories carousel. If your only goal was SEO, the equation has changed: you must weigh the real speed gains against the maintenance costs and tracking issues.

For media sites that rely on mobile traffic and need to be in news carousels, AMP remains relevant despite its flaws. For a corporate site or a small e-commerce site, a well-optimized mobile site without AMP is often more profitable and much easier to measure. Since Page Experience incorporates the Core Web Vitals, you can achieve excellent performance without the complexity of AMP.

How can you identify if your data is affected by this issue?

Start by comparing your metrics before and after AMP implementation. A sharp jump in mobile direct traffic simultaneous with a decline in mobile organic traffic, without any real change in awareness, is the classic signal. Specifically, look at the desktop/mobile ratio in each channel: if mobile makes up 60% of your organic desktop traffic but only 20% of mobile, that's suspicious.

Use behavior reports to see which landing pages dominate the direct channel. If they’re your standard AMP pages (blog posts, product sheets), rather than your homepage, you have proof that these visitors are not really direct accesses. True direct traffic generally concentrates on the homepage and a few strategic pages.

What immediate technical corrections should you make?

Correctly implement the domain linker in your amp-analytics configuration. The data-linker attribute with appropriate parameters allows you to transmit the client ID between domains. If you’re using Google Analytics, add the linker parameter to your tag with your main domain as the destination.

Set up automated UTM parameters on your AMP URLs. Create a rule to systematically add utm_source=google&utm_medium=amp-cache to all your AMP URLs served from the cache. It's a workaround that requires development, but it works. Just be careful not to break AMP validation with poorly encoded parameters.

Should you abandon AMP or can you live with this limitation?

Weigh the real benefits. If you’re in media or news, AMP remains an asset for visibility in carousels despite measurement issues. Just document the anomaly in your reporting so stakeholders understand that the apparent decline in mobile SEO is a tracking artifact.

If you’re in e-commerce or B2B with a long conversion cycle, seriously consider abandoning AMP. Speed gains can be achieved with good standard technical optimization (lazy loading, image optimization, minification), and you maintain clean attribution. The gamble often isn’t worth it anymore since Google loosened the eligibility criteria for mobile features.

• Audit your Analytics reports to spot a correlation between AMP implementation and increased mobile direct traffic
• Configure the domain linker (data-linker) in amp-analytics to transmit the client ID
• Add automated UTM parameters to AMP URLs to enforce proper attribution
• Create advanced Analytics segments to isolate and reattribute misclassified AMP traffic
• Document the anomaly in your dashboards and reports to avoid misinterpretations
• Periodically reassess the ROI of AMP: development maintenance versus actual SEO benefits