Three elements that drive SEM campaign performance in Google Analytics in the mobile traffic segment

Google Analytics is the most popular analytics system in the world, plus it has one undeniable advantage: with the free version I was able to successfully measure the third largest portal in Poland. In this article I will introduce three unusual elements that affect the results of SEM campaigns on Google in the segment of traffic from mobile devices.

1. SERP Bounce

That Google measures landing page quality is a fact stated in the official Google Ads documentation. This factor affects the quality score and ultimately the ranking of ads, i.e. the fact that we will pay more per click than our competitors if our website is weaker. The question is what exactly it means?

Well, one of the things that Google can measure is the reflections from the. For the sake of simplicity, we can compare the traffic sent to page A and page B and evaluate which result turned out better for users. If from page A as many as 50% returned to Google to look for other or similar results, and from page B only 10%, then we have a clear situation that on page B they were more likely to find an answer. The algorithm is much more complicated, but it all boils down to finding the sub-pages that users land on from Google that have very high rejection rates.

This is achieved by juxtaposing the dimension Landing Page, with the pair source/medium: google/organic. Then, based on keyword data from Google Search Console, we analyze what we can improve on the found landing page. This is the moment when ordinary SEO changes into SXO – Search Experience Optimization, and both a UX/UI specialist and necessarily a very good copywriter come into action.

2. Patching the holes, i.e. 404 alerts and server errors

Google officially says that if GoogleBot encounters too many server errors, it slows down indexing. To put it simply: if your website has errors, then new products and offers will appear more slowly on Google. Improving your system and detecting problems before they show up in Google Search Console is an important part of your SEO strategy. How Google Analytics can help? Simply place the code according to Google’s suggestions, i.e. on every subpage of the website, including the one that is responsible for the 404 Not Found error message or server errors e.g.: 503 Service Unavailable. Additionally, set up custom alerts, i.e. email notifications when the number of sessions on error-related pages exceeds the accepted values and voila, you will respond to errors faster than Google. Remember, these pages have different URLs, but you can give them a specific page title pattern, the TITLE tag in HTML is responsible for this, which is also sent to Google Analytics. It’s how you will know where and for what reason an error occurred.

You like order? Launch Majestic.com and in „pages” report you will know all landing pages which are linked to your website. Check their statuses, and in addition to errors, fix unnecessary redirection chains.

3. Page load speed report

At the end you will encounter the biggest disadvantage of the Google service, because this report is a trap of a small amount of data. Most often the data presented is based on a very small sample of sessions which took place on your website and cannot be the basis for further optimizations. It’s bad, because page load speed is an important factor affecting both the quality score in Google Ads and position in SEO efforts.

The most proven alternative here would be Google tools such as Lighthouse, built into your Google Chrome (F12 or Ctrl+Shift M and the horizontal menu on the right) and web.dev. Take into account that both tools test a particular URL, i.e. a particular subpage. Therefore, it is a good practice to scan many different sample subpage addresses, not forgetting such important ones as the sample search results page. Lighthouse additionally has a separate mode for mobile devices.

Interestingly, you can implement Core Web Vitals metrics in your Google Analytics using the method presented by Matteo Zambon.

LCP (Largest Contentful Paint) measures perceived page load time. Fast LCP will allow you to make sure that the page is usable. LCP should occur within 2.5 seconds from page load time.

FID (First Input Delay) measures interactivity. FID should be less than 100 milliseconds. This is another important and user-oriented metric – shows the usability of the website in terms of possible interaction from the user.

CLS (Cumulative Layout Shift) is responsible for visual stability, simplifying it determines how often a user may encounter an unstable and changing page layout. A good score is less than 0.1.

If you are interested in page speed, don’t forget about the Accelerated Mobile Pages (AMP) solutions available.

This article is part of a series in which we cover SEO, Ads, Analytics, and UX, focusing on three improvements in each area that are both scalable and have a big impact on the end result of a campaign.