Know exactly what users need
AI-powered user requirements and UX research support
Anyone who follows our projects knows that we attach great importance to good user research. If you develop a product without understanding the needs of the actual end users, you run the risk of missing the target group. User research is therefore indispensable, but unfortunately also demanding: it requires time, trained personnel and logistical effort.
What if there was software that made it possible to generate comprehensive proto-personas for different user groups at the push of a button? What if real user feedback from the field could be sprinkled in at any time and place to make the proto-personas more valid, step by step? And what if there was a product that could then summarize user requirements created in this way in an easy-to-understand and even entertaining way to efficiently pick up project stakeholders?
For LeanScope GmbH, we developed just such a software and brought in our many years of experience in human-centered development. We are proud to present an AI-based management tool that revolutionizes user research in all its facets.
AI-powered proto-persona generation.
At the heart of LeanScope is persona generation and its unique features. Using OpenAI's GPT and DALL-E 2, lifelike profiles of diverse users can be created. In addition to socio-demographic information, initial motivations and frustrations are also generated. Different roles can be assigned to the personas and they can even be related to each other. DALL-E 2 rounds out the profile with a unique photo of the person, without any copyright stumbling blocks. From here, the user research team can take over to sharpen the proto-persona step by step.
Unsure about a feature? Just ask! 🤷
Interactive chatting with personas in real time
A common problem: When developing a new feature, the question arises whether it is even needed by the desired target group. But time is short and it is difficult to quickly arrange an appointment with testers or real users. To be able to ask relevant human-centric questions at any time and from any place, we made it possible to chat with the personas. Thanks to AI, this feature enables a natural and quick exchange with the target group, be it for feedback, queries or even just a short poem about what annoys existing users the most *wink*.
What motivates my target group?
– Gamification & Player Types for Personas
Based on an article in the International Journal of Human-Computer Studieson the topic of gamification design, we have developed a test that can be used to quickly and easily determine a player type for the personas. This makes it possible to find out whether a persona is, for example, more of a communicative team player or has a competitive disposition. The playertype test improves the understanding of the target group's usage behavior, from which adjustments for the product can in turn be derived. In addition, it is also simply fun and provides surprising self-knowledge!
Our AI philosophy
When it comes to AI, we take neither a glorifying nor a judgmental stance. However, we are always open to new ideas and technologies that are fun and help us playfully achieve our goals and make our work more flexible and creative. For this reason, the topic of AI is already firmly integrated into the workflow at Centigrade. We are intensively engaged with it and are very aware of the possibilities as well as the limitations and dangers.
In line with our philosophy, we believe that humans must retain a validating and controlling role when working with AI. This position is reinforced by the fact that AI is still prone to error. Unfortunately, chatbots like ChatGPT tend to confidently present false statements - also known as "confidently wrong". So doubt is justified!
To get around this problem, we have configured the personas' AI to be somewhat reticent. Questionable or false statements are not simply made succinctly and confidently. Instead, when faced with a question for which it does not have a clear answer, the persona will indicate that it does not know for sure and recommend further manual user research.
In addition, user researchers can check the personas' characteristics, scenarios and needs against each other and flag any information they consider risky or worthy of verification.
Privacy is not just a nice-to-have
Data protection is an issue that is often neglected, and the rapid development of AI tools initially poses more security risks. There have already been cases of large international companies where several employees have worked with highly confidential data in ChatGPT, which could then be viewed by users around the world via the chatbot.
When developing LeanScope, it was therefore important to us that the users themselves can decide in a very fine-grained way which data the AI should have access to and which not. Not only does this make sense, but it's also required by law. Unfortunately, GDPR compliance is often just seen as a nice extra, rather than the guideline that it actually is.
You decide what the AI reads
This approach also distinguishes LeanScope from other AI-based tools. Often, the AI is given full rights and access to data that it doesn't really need - which is not appropriate according to the GDPR. To avoid cases like the one described earlier, it is important to consider applications like ChatGPT as "other employees", to whom the principle of least privilege applies, just like to any other employee.
Local servers and data sovereignty
For us, this also means that we don't want to leave it up to third parties where user data is stored. LeanScope's servers are located in Germany, so we have full control over what data is transferred to OpenAI's servers in the US, for example. This approach provides additional security for sensitive information.
Next Level User Research
Organising user requirements is a complex and small-scale process. Our goal with LeanScope was to centralise this process while keeping it simple and flexible. The continuous development of AI allows us to get more out of the tools than we could have imagined at the beginning, without losing sight of the value of humans in this process. We have been tasked by LeanScope to continue to implement new features quickly and continuously based on user feedback, even in the long term. We are proud to have played a major role in the design and development of such a unique tool and look forward to its further evolution!