The Emergence of 1st-Generation Recruitment Chatbots in Finland
Various digital and, more recently, AI-based tools have been introduced to help recruiters. These tools include so-called recruitment chatbots that aim to provide better customer service and candidate screening functions. We made a review of current chatbot applications as part of our “Recruitment Bots” project, funded by Intelligent Society profiling programme at Tampere University. Here are some picks from our findings!
The often-advertised premise of recruitment chatbots is that they could reduce the workload of recruiters by providing automatization to the search of candidates and the first contact point to the potential candidate. In addition, the bots are expected to reach and handle many more candidates than conventional, manual recruitment.
In Finland, there is a growing market space for companies that provide general chatbot services. Startups developing chatbots include GetJenny, Ultimate.Ai, Leadoo, ChatbotPack, Upseller, Giosg and Front.Ai. Some of the companies advertise the capability to understand written Finnish language and others rely on scripted conversations with predefined options.
Worldwide, there are numerous chatbot services that focus on recruitment. Probably the most famous examples are Mya, XOR, Olivia, Wade & Wendy and Ideal to name a few. These examples can often be found from recruitment bot related news hype articles and from startup seed funding news. However, it is still hard to find actual examples of people having been successfully hired with the help of such tools. Many of the services have existed already for years but there is no clear market leader.
In general, it seems that the chatbots are typically built on top of a web page or as a mobile application. The development of natural language processing (NLP) has enabled written communication that mimics human conversations. Chatbots are used broadly, for example, to assist customer service in the consumer market.
The first-generation recruitment-related chatbots could be classified into two categories. First, there are so-called lead bots that inquire and collect basic information about the job seekers by using scripted and relatively simple textual interactions. Second, there are customer service bots that provide answers to frequently asked questions about certain positions or the organization in general.
Technologically, it is noteworthy to stress that the lead bots do not have NLP capabilities but they rather work as conversational marketing technology. On the other hand, the customer service bots sometimes do make use of NLP (e.g. Seppo by GetJenny).
In Finland, a digital marketplace for jobs (fi: työmarkkinatori) has been under development for a couple of years and is currently undergoing beta tests. At some point, they were developing a chatbot that was meant to help job seekers to find relevant job ads based on their skills and interests. The predecessor, web site mol.fi publishes tens of thousands of job advertisements per month, so it could significantly increase the awareness of recruitment related chatbots. We are curious to see if and how the chatbot will be implemented in such a vast digital marketplace.
Also, research on recruitment chatbots has recently gained in popularity, and there are some studies specifically on recruitment chatbots. Regarding general chatbot research in Human-Computer Interaction, the work of Asbjørn Følstad [1][2][3]stands out. Regarding recruitment chatbots, Michelle X. Zhou have developed, together with her colleagues, Juji chatbots including “Empathetic AI Interviewer” that have been used in couple of studies. Chatbots and other forms of natural language user interfaces and conversational agents have several relevant publication forums such as ACM IUI (Intelligent User Interfaces), ACM CHI (Computer-Human Interaction) and recently started CUI (Conversational User Interfaces).
In our current research project, our objective is to study the appropriateness of recruitment chatbots, especially from the recruiters’ perspective. How do recruitment bots affect the decision-making process? Do the recruiters trust these tools and to the information they provide? To what extent the bots could be used to automate some of the decision making, mindful of the various social and ethical pitfalls? We are interested in the real-life experiences of recruiters, so if you have some to share, feel free to contact us!
