AI is rapidly evolving into a competitive advantage, prompting tech providers and organizations to integrate new features and enhancements. Google is investing heavily in AI-powered products, enhancements, and innovations. One such innovation is their new tool which is in Beta, Analytics Advisor, for which InfoTrust is proud to be an early tester and adopter. This post explores how to maximize the value of the Analytics Advisor and what you can do with this technological advancement.
What is GA4 Analytics Advisor?
Analytics Advisor is an AI-powered tool within Google Analytics designed to provide data-driven insights and recommendations. Looking and feeling like Gemini, it is much more advanced and intuitive than previous search bars in GA. It helps users quickly understand performance, investigate changes, and identify opportunities to optimize their strategies. More than just a tool, it acts as a thought partner within GA to maximize analysis value, offering visualizations and step-by-step instructions to turn insights into actionable improvements.
Overall findings and opportunities summarized
InfoTrust tested the new AI agent, Analytics Advisor, with customer-like queries across seven categories: general queries, general insights, advanced analysis, optimization, educational, administrative, and GMP integrations. Our general feedback is Analytics Advisor performed well on general analytics and optimization questions but still struggled with specific data analysis and custom dimension queries, offering high-level rather than property-specific insights. However, the tool is still in Beta, so we are hopeful Analytics Advisor will continue to improve and refine!
To get started with Analytics Advisor:
- Familiarize yourself: Ask Analytics Advisor to perform tasks like making a change in the GA4 UI or checking a data point to see how it works in your daily workflow. Here are a few prompts that the InfoTrust team found useful to get started (include link to the prompt blog post).
- Iterate with follow-up prompts: If the first prompt isn’t perfect, refine your query with a follow-up.
- Create a Shared Repository: Keep a shared document of your organization’s highest-value prompts.
- Respect Confidentiality: Do not enter sensitive or proprietary information. Check your company’s policy.
- Provide feedback to Google: Use the feedback button to report if a response doesn’t meet your expectations.
| Category | Types of Prompts |
|---|---|
| General |
|
| General Insights |
|
| Advanced Analysis |
|
| Optimization |
|
| Educational |
|
| Administrative |
|
| GMP Integrations |
|
Detailed Findings
General Queries
The analytics advisor performed really well against general queries such as “Where do I see attribution data?” and “What is the difference between total purchasers and purchases metric?” and either provided definitions or directions on how to view the desired report.
General Insights
The tool did a pretty good job of answering questions like “Has there been a downward traffic trend during any recent time periods?” and “How many purchases did I have for the month of September?” It provided a clear answer to the questions and even provided small, readable graphs directly within the chat interface.
Advanced Analysis
Specific Advanced Analysis questions like “Based on insights from Data Driven Attribution, what is the most effective channel path that is leading to higher conversions?” are where the tool seemed to stumble or not perform up to our expectations. It looks like there is limited advanced analysis that Analytics Advisor can do; however, it still gives some details that don’t directly address the question, but are helpful to consume.
Optimization
Optimization of data is another category our team felt the “Analytics Advisor” really seemed to shine. We tested questions like “Give me 5 recommendations to improve my website” and “What recommendations would you make for us to improve the key event? This is our main key event, and we would like to improve the conversion rate”. The tool provided actionable recommendations. Our only negative feedback on this topic is that the recommendations were a bit generic, and they didn’t actually link to the report mentioned. In one specific case, it mentioned the landing pages report as a recommendation, but the link provided was to the standard “Pages and Screens” Report instead. Not a huge inconvenience, but definitely could be improved on.
Educational
While the tool performed well against our various “Educational” queries, we feel that this could be one area where users will need to use caution. Like with most AI tools, “Analytics Advisor” pulls information from a wide range of sources related to GA, and those sources may not always contain the most accurate or up-to-date information.
Administrative
Analytics Advisor exceeded our expectations during administrative testing. Rather than simply providing navigation and instructions on how to complete a task, the tool elevated its support to act as a comprehensive administrative consultant.
For example, when we inquired about adding new users, the Advisor provided the necessary navigational steps alongside crucial insights:
- A breakdown of the different types of user permissions.
- Best practices for access governance and conducting regular permission audits.
- A holistic view of how the specific administrative configuration impacts the entire property or account.
This depth of information, covering configuration impact, best practices, and holistic functionality, turns a simple administrative request into a strategic action.
GMP Integrations
The performance of queries concerning integrations between GA4 and other Google Marketing Platform (GMP) products was generally strong.
The Analytics Advisor performed great when queries were focused on understanding GMP configurations such as providing details on the specific steps required to link your Property to a product in GMP, as well as outlining the benefits of that integration. One area of improvement is around specifics of the current GMP integrations in your account, and when focusing on these queries it was able to provide valuable insight in most cases, but the accuracy varies case-by-case.
Best Practice: We observed a higher success rate when the query was focused on a single product (e.g., “what Google Ads accounts are currently linked to my property”) rather than multiple products at once (e.g., “what Google Ads and BigQuery accounts are linked to my property”). This focus is key to obtaining reliable, accurate answers.
Conclusion
The AI-powered Analytics Advisor in Google Analytics shows significant potential for maximizing analysis value. While excellent for general queries, administrative tasks, and actionable recommendations, performance was inconsistent with highly specific advanced analysis and custom data dimension questions. As a Beta tool, further refinement is expected. The promising Analytics Advisor Beta will roll out to English-language accounts in December 2025, reflecting Google’s strong investment in AI analytical solutions.
