GEO, AEO and LLMO with etracker analytics

by Olaf Brandt

In the new era of AI-driven information discovery, online visibility is no longer just about traditional SEO, but about mastering the new disciplines of Generative Engine Optimization (GEO), Answer Engine Optimization (AEO) and Large Language Model Optimization (LLMO).

Ultimately, it is about being understood by the AI systems and appearing as prominently as possible in the responses. Similar to SEO (search engine optimization), web analytics can also provide valuable input for the new measures to be taken. This includes making the traffic from AI systems such as ChatGPT or Google AI Mode visible.

The former is particularly easy in etracker analytics. To do this, open the Medium marketing report and add the Origin dimension.

You will then find the AI medium, which you can expand to make the individual referrers or, in this case, LLMs visible.

The volume of traffic is usually still very low compared to SEO. However, the engagement and conversion rate may be noticeable. It is also advisable to drill down to the corresponding entry pages. This may allow conclusions to be drawn about the areas in which authority is assigned or the content preparation is beneficial.

Measure Google’s AI Overview Links

Google’s AI Overview (AIO for short) is a function that displays AI-generated answers and summaries above the conventional search results. Measuring how many users come to the website via this AI content is unfortunately not that easy.

The AI Overviews link to external content via the source reference. These links are often given the corresponding text fragments (#:~:text=) in the URLs. When users click on links in the summaries, they land directly on a very specific section of the linked page, which is then also highlighted.

These text fragments from the AI Overview links can be made visible in etracker analytics. This then looks as follows:

Here, too, it is advisable to evaluate the corresponding entry page.

These four steps are required to achieve this evaluation:

1. add a custom dimension of the type Page Impression via Account → Data enrichment/Custom dimensions.

2. create a variable in the etracker tag manager to read the text fragment from the URL parameter.

Insert the following code:

function testaio() {
  var nav = performance.getEntriesByType("navigation")[0];
  if (!nav) return;
  
  var hash = new URL(nav.name).hash;
  var match = hash.match(/:~:text=([^&]*)/);
  
  return match ? decodeURIComponent(match[1]) : undefined;
}

3. create a trigger that checks whether the AIO variable is filled.

Use the previously created variable as a user-defined event.

Regex for checking whether available: ^.+$

4. create a tag of the type etracker Dimension.

Use the previously created trigger and your own AIO dimension for this.

Optionally, you can now create a report configuration with your own AI dimension. To do this, simply call up the Mediums report, for example, add the AIO dimension and save it as a new configuration.