The new AI Deep Dives combine real, deterministically calculated SISTRIX data with the interprative strength of AI – and delivers the result directly in SISTRIX: concrete findings, clear action recommendations, no need for guessing. To begin with, seven analyses are available, from brand recognition to competitive fields and content opportunities.
Recently, we made the SISTRIX MCP Server, an interface in which AI systems such as ChatGPT or Claude can retrieve and categorise real SISTRIX data, even better with a new function usable for all – we go one step further. The combining of the expertise of Chatbot and SISTRIX data no longer takes place only in internal AI systems, but directly in SISTRIX products: with the new AI Deep Dives.
AI analyses combined with data & calculations by SISTRIX
A generic AI model without MCP connection sees a table and can describe it. AI is good at reading and interpreting large datasets quickly. However, it is not good at delivering reliable calculations; with a thin baseline it invents facts which means that distinguishing trustworthy expert knowledge from dubious claims is almost impossible, since both sound equally plausible.
This is why we separate these tasks for our AI Deep Dives: the AI interprets – the data, deterministically calculated numbers and the verified SEO – and the GEO knowledge comes from SISTRIX.
This ensures that the result has no invented numbers but instead substantiated findings that not only state what the datas hows but also what follows and what should be done. The expertise from SEOs is not being replaced, but made available more quickly and the evaluations are comprehensive without deeper SEO or GEO knowledge.
The AI Deep Dives can be found where you need them: in the filter bar of the corresponding function, visually highlighted in purple from the regular filter buttons.
7 Deep Dives in the AI/Chatbot Beta
The starting point is made up of seven AI Deep Dives in the SISTRIX AI/Chatbot Beta that we will briefly present in the following. We are in the process of creating detailed tutorials for special features such as Use Cases or the different skills for you.
In the AI Check, the area, in which you can enter a domain or brand and see for which prompts it is mentioned or cited in AI responses, the Deep Dives are available for these features:
- Prompts
Brand perception: Is the brand featured in a positive, neutral, or critical light? Which strengths and weaknesses are mentioned and how does this differ between platforms? 
Fact consistency: Is the information consistent across different AI platforms? Which claims vary and where do models even contradict each other?
Topics & Roles: For which search request, apart from brand searches, is the brand named? ANd in which role, as a central recommendation or only on the side?- Sources
Own sources: How much of the cited content is coming from your own sources, which pages are used more frequently and which directory collects the most citations? 
Outside sources: How much content is from outside sources, which pages are the most cited third party sources and which type of content brings the most mentions?- Competition
Competitive environment: Which companies are named in many prompts as well, which brand is the biggest competitor in the LLMs and how are competitors positioned as compared to your own brand?
Source gaps relating to the competition: In which content areas are competitors cited but your own brand is not? This shows the concrete gaps that can be closed.
The different AI Deep Dives deliver data-based analyses directly in the familiar SISTRIX interface. Additionally, the created analyses can be downloaded as a markdown file or PDF.
Free for all SISTRIX users
The AI Deep Dives are, for now, available in SISTRIX for AI/Chatbots – free for all SISTRIX users. We will initially collect feedback and experiences for the different features. To add your feedback, there is a corresponding feedback field under every AI Deep Dive.
We appreciate any feedback you let us know there (or directly sent to support@sistrix.com) – positive and negative, so we can enhance these features and offer them in other areas in SISTRIX as well.
Have fun trying it out!
