Our AI assistant analyses your UX tests and interviews
The AI assistant is currently in the beta phase. Sign up here and help us to further develop the AI analysis.
Analysing your UX research can be quite time-consuming: many hours of watching videos, writing detailed notes, creating summaries and reports. Meanwhile, the next research projects are already lurking in the background.
The benefits of using RapidUsertests AI:
- Less effort: You save a lot of time analysing the tests and can process a lot more research tickets
- Increase impact: You will be able to communicate insights within your company much faster and thus increase the UX department's influence
- Time for bigger projects: This leaves more time for strategic/fundamental research
- Forget boring analyses: AI takes repetitive work off your hands
- Involve other teams: AI is intuitive to use and removes the barriers for beginners to get started with testing. This means you can introduce other teams such as marketing and design to user testing
How accurate is the AI?
The AI assistant analyses your tests at video level and documents the findings. It identifies 85-90 % of insights obtained by our human researchers.
How did we achieve this? We have been working on analysing UX tests and research using AI for over a year now. To do this, we utilise the collective power of our UX researchers and constantly compare the results of the AI with their own results.
In addition, we continuously gather feedback from our customers during the beta phase. This enables us to develop an assistant that is completely geared towards the needs of its users.
So far, the AI assistant is still limited to unmoderated tests. However, we are working at full speed to teach it to analyse recordings of moderated tests and interviews.
What is the AI assistant capable of?
The AI assistant prioritises the findings according to their severity and distinguishes between positive and negative comments.
It also uses project-based tags that it derives from the test concept automatically. This gives you a quick overview of the findings and the most important topics.
You can transparently track which comments were generated by the AI and stay in control of your documentation at all times.
How does this work in the tool?
- Once your account has been activated for the trial period, you can activate the AI analysis in your dashboard on the test overview page.
- It only takes a few minutes for the insights to be available.
- We recommend that you first get an overview of the insights in the ‘Your analysis’ section. Here you can filter by tags and priority, and jump directly to the relevant video clips.
- You can easily add further findings yourself and edit the comments generated by the AI.
How can I be sure that the AI is not ‘fantasising’?
Our AI assistant is based on generative AI i.e. a large language model. Many of these AIs were created to generate new output, which means that they may come up with imaginary information. This is not the case with our AI. The RapidUsertests AI assistant is set up so that it draws its statements exclusively from the transcripts of the test videos and does not ‘fantasise’ any additional information.
To be extra sure, you can double-check each statement yourself by watching the linked video passage.
What about data protection? How secure is the data?
We know how important your data is for you, your company and your test subjects.
That's why the AI assistant is in compliance with the GDPR which is ensured by our TÜV and DEKRA-certified data protection consultant.
As always, the test subjects take part in the tests with a pseudonym and also agree to the transfer and saving of their personal data.
You can find out more about data protection at RapidUsertests here.
Where is the data being processed?
We use our own instance of Microsoft Azure to process the data. The transcripts of the videos are stored on servers in Frankfurt am Main. The analysis takes place on servers within the EU.
Will my data be used to train the AI?
All data processed by the AI assistant is protected and is not used to train third-party AI models.
Sounds good? Even though we already are very proud of the AI assistant, we still see a lot of potential in how it can best support you in your work. That's why we need your help to develop it further! Sign up for our beta phase and help us to further improve the AI Assistant.