AI overviews, chatbots and more: SISTRIX updates in the first half of 2025

New features for AI Overviews, a visibility index for chatbots, targeted analyses in generative systems and many minor improvements – SISTRIX continues to evolve with search. Here is an overview of the most important innovations from the first half of 2025.

We are currently experiencing a transformation in the SEO industry unlike anything we’ve seen in a long time. Innovations such as AI overviews in Google Search, AI Mode (already launched in the US), AI search engines and chatbots are changing how and where “search” takes place today and in the future.

Against this backdrop, the first half of 2025 at SISTRIX was also dominated by changing search habits. Because one thing is clear: if you want to be successful in the new search era, you need reliable data.

Tools such as SISTRIX are essential for maintaining an overview in this dynamic landscape and making the right strategic decisions.

Despite all these changes – the majority of searches still take place via Google. That’s why we are continuing to work on our core functions and striving to become even better in this area too.

In this article, we summarise all the changes made to our products in the first half of 2025:

AI Overviews: new analysis functions in SISTRIX

AI Overviews are Google’s response to the surprising success and growing market share of ChatGPT, and at the same time the first step towards integrating AI functions directly into the familiar search results list. Since users do not have to (or cannot) actively opt in, they are currently the most widely used AI feature in web search.

Finally rolled out in the UK at the end of 2024, generative AI responses now appear in over 15 percent of the search queries.

AI overview for the keyword "what is a flat white" with structured information on preparation, taste and origin.

In terms of content, AI Overviews are similar to the well-known Featured Snippet and pursue the same goal, namely to answer simple user questions directly on the Google results page. Unlike the Featured Snippet, however, Google does not rely on a single source, but combines content from pages that already rank organically for this keyword, usually in the top 10.

However, this presents website operators with a number of new challenges. In particular, the lack of insight into impressions and clicks generated (or not generated) by AI Overviews in Google Search Console makes analysis difficult. To close this gap, we have integrated three new features into SISTRIX that allow you to analyse which keywords AI Overviews are displayed for and which websites are named as sources in them.

  • A new keyword filter called ‘AI Overview – with ranking’ allows you to filter specifically for keywords for which Google displays an AI Overview. Using a ‘with ranking’ checkbox, the filter can also be expanded so that only those keywords remain for which the analysed domain is named as the source within the generated response.
SISTRIX keyword view for lego.com with active filter "AI Overview (With Ranking)" and SERP feature menu open.
  • In addition, the new menu item “AI Overviews” can now be found in the left navigation menu under SERPS. This overview page lists all keywords for which there is an AI Overview in which the entered domain is cited. Useful details are displayed for each keyword: Which URL of the domain is cited in the overview and at which position within the answer box the mention occurs.
  • Finally, we have expanded the SERP archive: For each keyword, the complete content of the AI Overview can now be displayed, including the generated text and all source references (linked URLs). SISTRIX even recognises named entities in the response text to better understand which people, brands or concepts are mentioned.
SISTRIX Mobile SERP with AI overview for the keyword "flat white" and brief explanations of the ingredients and preparation of the cocktail.

These new features help to better understand the impact of AI overviews on organic clicks and rankings. You can now see at a glance which search queries Google considers a domain in its generative answers – and how prominently it is placed within the AI results. A detailed explanation of the new features for evaluating AI overviews can be found in our changelog for AI overviews.

The AI Overview analysis functions are now available to all users. If you do not yet have access, you can create a free trial account at any time.

The next step – Google AI Mode

With AI Mode, Google is currently testing a new form of web search in the US that combines classic results with a dialogue-based interface. The feature is still optional and only available in the US, but a wider rollout, similar to AI Overviews, is considered likely.

Since impressions and clicks from AI Mode are not currently reported separately in Google Search Console, there is a lack of reliable data on usage. We are already working on making this new search mode evaluable.

Google search in AI mode with AI response to déjà vu.
Source: blog.google

If you would like to receive regular updates on the status of AI Mode, new SISTRIX features or other developments related to SEO and visibility, you can subscribe to our monthly review newsletter, in which SEO expert Jono Alderson summarises all the important news in a concise format.

Understanding visibility in chatbots: The SISTRIX AI analysis

In addition to Google, independent AI systems such as ChatGPT, Gemini and Deepseek are also constantly evolving. They no longer just answer emails or provide translations, but are increasingly answering knowledge-based questions, thereby shaping digital search behaviour. ChatGPT in particular is now one of the most visited websites in the world.

Some of these systems combine classic web search with AI models: a search index provides relevant websites, and the language model reads the content and formulates its own response. This is reminiscent of Google’s AI Overviews, but is usually displayed in a dialogue-like interface. Interestingly, both Perplexity and ChatGPT use Bing’s index for their web searches.

Local restaurant search in ChatGPT with interactive map for Bonn and two Michelin-listed restaurants, including opening hours and links.

Unlike Google Search, however, chatbots rarely cite direct sources. The information often remains opaque and is also based on outdated data, as new models are only released at irregular intervals. At the same time, responses can vary greatly depending on the prompt. This dynamic has made reliable analysis virtually impossible so far.

This is exactly where our new product SISTRIX for AI/chatbots comes in.

SISTRIX AI overview for the topic "Flat White" with visibility values in various chatbots, brand ranking and prompt examples.

We carefully select 10 million prompts to ensure that they represent a cross-section of actual chatbot usage in the respective countries and languages. These prompts come from various sources. These include, for example, prompt suggestions, frequently asked user questions from Google search results, and other relevant data sources.

The data situation and the availability of reliable foundations for LLM/AI analyses are currently even more challenging than, for example, at Google or Amazon. However, this is a development that we are familiar with from the early days of Google analysis. There, too, data quality has gradually improved. We are confident that this will also be the case with LLM data.

Entity environment for "cappuccino" in SISTRIX with related terms such as espresso, milk and milk foam.

We currently support OpenAI ChatGPT, Google Gemini, and DeepSeek. If new, important market players emerge, we plan to include them as well.

In the responses from LLMs, we recognise entities such as brands, people, places and more, as well as the associated source references. Similar to Google and Amazon, this creates the possibility of reverse searching: Which prompts mention a brand or its competitors? How visible is a brand?

SISTRIX for AI/chatbots is currently still in beta version and is being continuously expanded. Activation is currently still done manually, but we expect to be able to grant all users access to a finished product soon.

The new AI tool in SISTRIX provides, for the first time, a reliable database on the visibility of brands in chatbots such as ChatGPT, Gemini, and Perplexity – including specific prompts, competitors, and environmental analysis. Try SISTRIX free for 14 days and find out how your brand performs in the most important AI systems.

Further innovations in the second quarter of 2025

In addition to the major updates, there were a few minor improvements in Q2, which we will briefly summarise here:

Keyword ideas – Questions

In the keyword ideas section, the “Questions” tab displays Google search queries that contain both the selected keyword and typical question words such as what, how, when, where, etc. This often makes it easier to understand the search intention behind a keyword.
A new feature is that SISTRIX now also displays the number of question search queries found. This number describes the sum of all keywords that can be assigned to the start keyword as a specific question.

SISTRIX page with frequently asked questions about the keyword "cappuccino", including search volume and wording variations.

Segment filter for AI overviews

Segment filters can be used to analyse visibility for specific areas of a domain, such as individual brands, topics or product categories. Previously, such individually saved filters could already be used in visibility analysis to highlight specific developments.

A new feature is the combination with the keyword filter “AI Overview – with ranking“. This allows you to create a segment that only contains keywords with AI overviews and view it directly in the visibility history.

This is particularly helpful if you want to analyse how visible a domain is in certain areas of generative AI search – for example, for non-branded keywords or individual product lines. The combination of both filters makes this possible.

For more information on how to set up the filter and how it works, please refer to the changelog for segment filters.

Review: New features from the first quarter of 2025

There were also numerous innovations at SISTRIX in the first three months of the year. Here is a concise overview of the features from the first quarter:

Amazon reviews summarised: SISTRIX’s Amazon analysis captures AI-based summaries of product reviews. They show the most common pros and cons of a product without having to read all the individual reviews. This makes it easier to identify customer trends and evaluate reviews more efficiently. (More on this)

AI analysis of UGREEN webcam: Image quality is rated predominantly positive, with 54 positive reviews to 11 negative reviews.

On-page optimisation: The on-page projects in SISTRIX have been expanded. Changes to a URL can now be crawled immediately without having to update the entire project. This saves time and helps to check optimisations directly – e.g. after adjustments to loading time, meta data or structured data. (More on this)

SISTRIX URL Explorer with details about the checked page, status code 403, crawl time and the option to restart the crawler.

API updates: At the beginning of the year, we completely revised our API documentation. Queries can now be tested directly in the browser, using real or demo data. Several new functions have also been added:

These new features make it easier to integrate SISTRIX data into your own workflows. (More on this)

SISTRIX API documentation with an overview of functions, endpoints and access via API key.

Labs features: In SISTRIX Labs, we regularly provide new or experimental features that can be activated manually. They are aimed at anyone who wants to test new features early on. Some features make it into the product later, while others remain optional extensions. New features include:

  • Global data: Enables cross-country analysis of visibility. Particularly helpful for international SEO strategies.
  • Search volume forecast: Shows a statistical estimate for the development of a keyword over the next twelve months based on historical data.
  • CTR detailed analysis: Compares the expected click-through rate per keyword with the actual values from the Google Search Console – including snippet preview and benchmarking. Ideal for identifying optimisation potential in SERP displays. (More on this)
Beta features activated in SISTRIX Labs for CTR Potential, Search Volume Forecast and Worldwide Data.

The world of SEO is evolving – and so is SISTRIX

The SEO landscape is changing. Google is testing new AI features such as AI Overviews, chatbots are becoming a real alternative to traditional search, and platforms such as Perplexity and Google’s AI Mode are showing where the journey is headed. At the same time, much remains the same – even in 2025, most search queries will still run through the traditional Google ranking.

For SEO, this means one thing above all else: if you want to remain visible, you need good data. Whether it’s classic search results or presence in new AI systems such as ChatGPT, Gemini or Deepseek – with SISTRIX, both areas can be reliably analysed.

The new features are now available in your account and can be tried out immediately. As always, we welcome feedback and suggestions (support@sistrix.com) so that SISTRIX can offer exactly the features you really need in your everyday work. Have fun trying them out!

Related posts