In the rapidly evolving landscape of digital marketing, mastering the intricacies of AI is no longer just an advantage; it’s a necessity. In our previous article, we explored the strategies that can amplify your content’s visibility in AI-centric search results. But what comes next? Developing high-quality content is only half the battle—measuring its effectiveness is vital to ensuring both value and content creation direction. How can you determine if your optimizations are truly making an impact and driving value? In this article, we’ll leverage Google Analytics and unveil actionable insights and reports that will help you track and evaluate the performance of AI-driven traffic.
Business Model Most Affected
To start off, we must first understand what business models are more heavily affected by AI overviews and reasons for why that might be the case. As of today, a majority of the traffic that is being disrupted by AI is organic search traffic due to the nature of the queries. While paid traffic is also still affected at a high rate when AI Overviews (AIO) are present on the pages, organic search traffic resulted in roughly a 70 percent decline in CTR when an AI Overviews is present. The reason for this has to do with the fact that AI searches typically involve +4 words used in the query and roughly +95 percent of Google organic searches are +4 words. In fact, since LLMs were released, the size of search queries only seem to continue to increase in size.
What this leads us to conclude is that previously high-performing content targeting organic search clicks on long search queries have been substantially affected by AI and AIOs. These business models tend to be heavily reliant on blogs and articles to drive traffic, as these content pieces are generally used to target organic search queries through keyword stuffing. Thus, if your organization’s business model includes leveraging large amounts of content to driving sales and transactions (B2B Services), it is likely that traffic will be more affected than business models that do not (B2C Retail).
Just to clarify, even though your organization’s business model may be affected due to AI being more heavily leveraged, it is important to continue to develop and post content to your site. However, instead of solely focusing on historical SEO practices, it is important to combine them with AI SEO best practices as this will help increase your contents’ visibility in AI search engines and traffic being directed to your site from AI. For more details on how to optimize your content for AI SEO, check out the first article in this Mastering AI For Marketing series.
I have said it before, and I will say it here again: In this ever-evolving world of AI it isn’t a question of how do we get back to what we used to do as an organization, but how do we gain a competitive edge over our competitors. One way to start this journey is to understand what AI search engines are available and leveraged by users so we can set up Google Analytics efficiently to analyze that traffic.
AI Search Engines
Understanding which AI search engines users might be leveraging the most and what kind of signals (UTMs, referrers, etc.) are sent when users select links that direct them from the AI search engine to the site is a vital piece we need before we can start analyzing traffic. A great place to start is through a generic search to identify the most popular AI search platforms that are currently available to the market based on a number of lists. Surprisingly, a majority of these lists had the same five to six AI search engines listed with a few other unique AI search engines also included. While many have heard of the top two or three AI search engines (Perplexity, ChatGPT, and Gemini), there are some that may be new to folks. That being said, the top search engines that we evaluated based on the lists above were:
- Perplexity : An AI-powered search engine delivering direct answers to queries using natural language processing for context and concise information. If you are interested in learning more about Perplexity, check out this article InfoTrust published that dives deeper into the unique dynamics this platform offers.
- ChatGPT by OpenAI : A conversational AI model designed for interactive text exchanges, providing answers, explanations, and engaging discussions on various topics.
- Gemini by Google : Combining large language models and advanced reasoning for dynamic and context-aware interactions.
- Chrome AI Overview by Google : Enhancing search algorithms and natural language processing for improved user experiences on Chrome web searches.
- Microsoft Copilot : Integrates AI in Microsoft Office applications to assist users with tasks through content generation, summarization, and data insights.
- Bing AI Overview by Microsoft : Integrates AI in Microsoft’s Bing search engine to provide relevant results and improved natural language understanding through machine learning.
- Bing AI Overview – Deep Search by Microsoft : Focuses on comprehensive analysis and synthesis of information for complex queries, delivering holistic insights.
- Phind : A search engine tailored for developers, providing coding information, snippets, and documentation for tech-savvy users.
- Andi : An AI search engine presenting information in a user-friendly format with an emphasis on privacy, returning conversational responses.
- You.com : A customizable, privacy-focused search engine allowing users to personalize results and aggregate information from selected sources.
- Exa : An AI-powered search engine focused on direct answers and synthesized insights across topics to enhance the search experience.
- iAsk : A question-and-answer-based search engine providing fast, accurate responses generated through AI analysis of user queries.
- Felo.ai : Delivers direct answers and insights through conversational interactions using AI.
Key Traffic Parameters
Now with the list of AI search engines that we will be leveraging and evaluating in this article, we must turn our attention to what data signals are passed when a user clicks on a link in the AI search engine and are directed to our site. In Google Analytics, there are three primary parameters that are important to ensure traffic is attributed effectively for reporting purposes. Those parameters are:
- Source : Identifies where website traffic originates, such as specific platforms or domains (e.g., google.com, newsletter). In Google Analytics, this information helps marketers evaluate the effectiveness of various traffic sources, enabling data-driven decisions to optimize marketing strategies and campaigns.
- Medium : Categorizes the type of traffic channel, such as organic, CPC, referral, or e-mail. Google Analytics allows marketers to analyze the performance of different channels, helping them understand which methods drive the most engagement and conversions.
- Referrer : Specifies the exact URL or domain from which visitors came before reaching your site. In Google Analytics, this insight helps identify specific webpages driving traffic, allowing marketers to assess the impact of partnerships, guest posts, and online ads, thus informing their optimization efforts. The value found here is primarily utilized to inform the value that can be found in the source as well.
Based on the list of most-leveraged AI search engines and the key traffic parameters leveraged for attribution purposes, we can start mapping what the expected values we would see in Google Analytics to the AI search engine for analyzing AI traffic. Below is a table of values that are expected to be observed when specific AI search engines direct users to your organization’s site:
| AI Search Engine | Source | Medium | Referrer | GA Default Channel Grouping |
|---|---|---|---|---|
| Perplexity | perplexity.ai | referral | https://www.perplexity.ai/ | Referral |
| ChatGPT | chatgpt.com | referral | https://chatgpt.com/ | Referral |
| Gemini | gemini.google.com | referral | https://gemini.google.com/ | Referral |
| Chrome AI Overview | organic | https://www.google.com/ | Organic Search | |
| Microsoft Copilot | (direct) | (none) | None | Direct |
| Bing AI Overview | bing | organic | https://www.bing.com/ | Organic Search |
| Bing AI Overview - Deep Search | bing | organic | https://www.bing.com/ | Organic Search |
| Phind | (direct) | (none) | None | Direct |
| Andi | (direct) | (none) | None | Direct |
| You.com | (direct) | (none) | None | Direct |
| Exa | exa.ai | referral | https://exa.ai/ | Referral |
| iAsk | iask.ai | referral | https://iask.ai/ | Referral |
As seen above, many of these AI search engines don’t follow a consistent attribution pattern because of the referrer. In fact, only ChatGPT utilized manual UTM tagging in addition to the referrer being passed just as a safeguard to help protect attribution. Many of the AI search overview features (Chrome and Bing) baked into the search engine results page (SERPs), ended up as organic search traffic, while some (Microsoft Copilot and others) ended up as direct traffic. While there is no real solution to help solve for the traffic that falls into these two buckets, there is an opportunity to leverage the other AI search engine traffic (ChatGPT, Gemini, Perplexity, etc.) for more analysis capabilities.
Note: While evaluating traffic to our site, it was noticed that there are also some other AI search engines that also directed traffic to content on the site. It could be important to evaluate your organization’s traffic sources to observe if there are other values that should be included in the below. The AI search engines that have hit our site, but are not listed above are felo.ai, blackbox.ai, and blurbs.ai. It is likely that a yearly review of traffic could be helpful to catch nuances like these.
Configuring Google Analytics
With an understanding of what AI search engines may be sending traffic to your site and where the traffic from the platforms is being attributed based on the traffic parameters, configuring Google Analytics (GA) for analyzing traffic can now be achieved. The first step to configuring GA is to set up custom channel groupings that bucket these identifiable AI search engine traffic. To complete this:
- Navigate to the GA property you wish to adjust and access the property admin section. Once in the admin section, navigate to the channel group setting.
2. Select “Create new channel group” to create a custom channel grouping. If your organization has a Custom Channel Group already set up that you wish to add this new AI channel grouping to, select the channel group instead of the button.
- Note: The Default Channel Group is not able to be edited or updated at this time.
3. Select the “Add new channel” button to create a new channel grouping to bucket AI search engine traffic together.
4. Give the channel a name and add a channel condition. Once users are satisfied, they may save the channel. The custom channel grouping feature is retroactive, thus channel conditions can be updated/adjusted as needed. The channel condition utilized to create the new AI search engine traffic is as seen below:
- Source matches regex .*[Pp]erplexity(?:\.ai)?|.*[Cc]hatgpt(?:\.com)?|.*[Gg]emini(?:\.google\.com)?|.*exa\.ai.*|.*iask\.ai.*
- The above Regex condition evaluates to ensure that the major AI search engines include values that may leave off the top-level domain (.com or .ai) and capitalized the first letter of the name, if they were ever to pass source values as a UTM.
- Medium matches regex referral|\(not set\)
- There are some instances where AI-drive traffic may not include a Medium value. In this case the Medium value will be “(not set)”. This is an example as to why it is important to evaluate the AI-driven traffic being sent to your site.
5. The last adjustment necessary is to reorder the channel grouping to ensure that the newly created AI search engine channel is ordered ahead/above the Referral channel. Because GA evaluates the channel matching conditions based on order, if the newly created channel is not ordered ahead/above the Referral channel, the AI search engine traffic will continue to be allocated to the Referral channel. Thus, select the “Reorder” button and drag the newly created channel ahead/above the Referral channel.
6. Once complete, select the “Apply” button and save the custom channel groups.
To ensure that the channel group is working properly, navigate to one of the acquisition reports (User or Traffic) and apply the custom channel grouping to the report. With the custom channel grouping, evaluating AI-driven traffic to your site will be substantially easier and more readily available for use by other in the organization.
How to Analyze AI-driven Traffic in Google Analytics
With GA now configured properly to help speed up the process and bring awareness to AI-driven traffic, organizations can start to compare data points across channels. There should be two primary goals that are achieved through analyzing this kind of data: how valuable is AI traffic to your organization’s business outcomes today and in the future, and what pages are driving the most AI traffic interaction to my site.
When analyzing AI traffic to determine how important it is to your organization’s business outcomes, leveraging the AI channel grouping and comparing it to other channels’ metrics can be tremendously helpful. Specifically comparing key event rates and engagement rates can help organizations determine if the traffic coming from AI platforms is performing well. While it is expected that AI-driven traffic is not going to be a leading channel when comparing users and sessions, evaluating trends via line charts can be helpful to understand both current performance and which channels may be being affected by AI. It is more than likely that organizations will see AI traffic increasing over time and channels like paid and organic search decreasing (also note that AI overviews directed traffic is also included in these totals).
Once the value of the AI traffic is determined, your organization will have a better understanding of how many resources to commit towards editing/creating content that may be more heavily leveraged by AI. One strong starting point towards understanding what content could increase AI-driven traffic is to understand what pages currently on the site are responsible for the most traffic. Typically, these pages are going to be related to trends, comparisons, lists, and technical how-to guides as these usually involve longer search queries, but they may not be limited to these. Once top-performing pages are identified, analyzing content structure and subject categories can inform your AI content strategy.
While many organizations are in the early stages of understanding and leveraging AI, analyzing the effects of AI on business outcomes is in an even more nascent stage. Taking practical steps today can not only lead your organization towards a competitive advantage, but can also save many headaches in the long run and keep your organization from falling behind.
Top 5 Recommendations for Analyzing AI-driven Traffic in Google Analytics
- Adopt AI-Enhanced SEO Practices : Transition from traditional SEO methods to incorporating AI SEO best practices. This will help improve content visibility in AI search engines and drive traffic that may have declined due to AI Overviews.
- Monitor and Analyze AI Traffic : Utilize GA to set up custom channel groupings specifically for AI-driven traffic. Focus on key traffic parameters such as Source, Medium, and Referrer to better understand the impact of AI on your website traffic.
- Identify High-Performing Content : Analyze existing content to determine which pages are generating the most AI-driven traffic. Target content related to trends, comparisons, and how-to guides, which are likely to attract longer search queries and engagement.
- Stay Updated on AI Search Engines : Regularly assess which AI search engines (e.g., Perplexity, ChatGPT, Gemini) are being leveraged by users. Continuously evaluate new AI platforms that may emerge to adjust your marketing strategies accordingly.
- Focus on Competitive Advantages : Shift organizational focus from reverting to past strategies to identifying ways to gain a competitive edge in an AI-driven market. Investing in understanding AI’s impact on user behavior and preferences will support long-term growth and sustainability.
Conclusion
The rise of AI is changing digital marketing and impacting businesses that rely on organic search traffic. To stay ahead, businesses need to adopt new content strategies that use AI-driven approaches to improve visibility and engagement. By using data insights from AI search engines and keeping an eye on traffic patterns, organizations can adjust their content strategies to meet the needs of today’s users. Adapting to these changes will not only help businesses overcome the challenges posed by AI but also open up new opportunities for competitive advantage. In the ever-changing digital landscape, being proactive and adaptable will be the key to success in an increasingly AI-driven world.
