Let’s face the truth: when it comes to search for information and services provided by the public administration, most people spend quite a lot of time surfing on the Internet, searching for keywords on the municipality’s website, going back and forth, to end up frustrated and giving up. The last resort: calling the public administration contact center in the hope that someone on the other end of the line can help them out. Why? Long story short, in most of the cases, people have scarce domain knowledge, bureaucratic jargon often sounds gobbledygook and, just to make things more difficult, website interfaces of public administrations are usually terribly complicated and cumbersome to navigate.
So what? Hence the idea of developing a chatbot that pursues exactly this objective: making information and smart city services easily accessible and usable by citizens, literally “speaking their own language”. Indeed, chatbots enable people to interact with a company or, as in this case, with the public administration in ways that are very similar to two-way human conversations. They are available 24/7 on a smartphone, making them perfectly positioned to provide the sort of anytime required information. Moreover, chatbots have been recognized as a tool able to serve customers in a meaningful way and, more importantly, at a substantially reduced cost (compared, for instance, to other types of customer service tools, such as phone assistance).
The occasion to translate this idea into something concrete started at the beginning of this year when the SynchroniCity consortium selected U-Hopper as one of the open call winners to pilot an AI-enabled service with the aim to improve the lives of European citizens.
We already talked a little about this project (link/general) and experience in two previous articles on our blog; however, this time we want to share with you the main learnings we take home from this 6-month journey. In fact, at U-Hopper, we believe every project is a new opportunity to experience, test and pivot our ideas and hypothesis. And this time was no exception!
While analyzing gathered data on the usage of our solution developed for the citizens of Milan, we noticed that many foreign citizens interacted with the chatbot asking for information and typing in questions in English and Arabic. We started wondering why our chatbot gained so much popularity among non-Italian speakers, proving to be a target segment with a greater aptitude for using it. We have always seen chatbots as a mere means for facilitating access to information and services, but we have never considered the potential of chatbots as a tool for driving integration and social inclusion; and this was for us an unexpected but pleasant surprise!
Though, if we think about it, a chatbot embedded into existing platforms such as Facebook Messenger (like ours) involves really low (or even not at all) barriers in terms of adoption, since users are already familiar with the interface - in the end, it seems like chatting with a friend! This represents a great advantage over applications and other ad-hoc means where users have to first “learn” the tool before actually using it. For foreign people, this is a great incentive: they know how to use Messenger and the perceived friendliness of a chatbot can help raise their comfort level in making questions while not feeling embarrassed, and without having to physically visit an office or talking with a person in a different language from their mother tongue.
Our chatbot was unfortunately Italian-speaking only, and we did not manage to provide full support to this large slice of our users. We are sure that if we had developed an English chatbot enabled by an NLP model, we would have succeeded in making information and services not only accessible and usable, but also relatable and explicable - with the (positive) side effect of increasing the capacity of the local authority to take concrete actions towards the integration of minor foreign communities.
Apart from empowering our technical skills, the most significant outcome of this project consisted, indeed, in proving that the inclusion of foreign people can (also) pass through the digital world and, more specifically, through a tool that is extremely easy and immediate to use! The potential is already there, what we need to do is to strengthen the culture of chatbots in our societies and to see the great results it can bring along!