“Will this be a dealbreaker?”

We’ve noticed that candidates approach recruiting chatbots more than half of the time outside of the working hours; so, what topics do they ask about?

In an analysis from a few months ago, we saw some broad topics:

• 25% of end-user queries fall into 5 major themes: Application Process; Salary; Professional Growth & Development; Internships; Contact a Human
• 25% of end-user queries fall into 14 minor themes: Application Evaluation; Application Deadline; Application Delete or Modify; How Long to Apply and Hear Back; Qualification; Application Documents; Language Expectations; Thesis; Working Hours; Location; Starting at the Company; Commute; Equipment; Benefits.
• 40% concern company-specific FAQs, and the rest are too specific (e.g., contain identifying information), or are noise.

There are some fascinating use cases in the last 40%: like responding quickly to emerging topics. For example, a candidate may have received a fraudulent phishing email claiming to be from a well-known employer brand; the candidate can go to the chatbot and ask, “I got an email from xyz@abc.com about a position, is this real?” — and through a machine-learning mediated layer, the recruiters can quickly create an automated response.

But even within commonly-occurring topics, directness is an interesting theme. When interacting with a conversational agent, people get right to the point: one of the most common questions across systems is “what is the salary?” — as well as questions about benefits, parental leave, and so forth. In particular, candidates ask about dealbreakers: “I have tattoos; will it be a problem?” or “My English is not very good, can I still apply?”

These are particularly interesting queries because they focus on skill or requirement lack, in a very straightforward way, often without hedging. The answers given by recruiters through the conversational agents are also more direct, at least by virtue of chat messages being relatively short.

So, why is it easier for people to pose direct questions to a bot and not to another person? One hypothesis could be that the conversational agent here is more human because it is not human: if you are interested in a job, you might hesitate to just ask outright about something that you already think might be a dealbreaker. However, there is no penalty to asking an automated, anonymous system. And in the best-case scenario? The chatbot delivers an answer, written by a recruiter, that anticipates this potentially complicated topic, and provides information that might help the candidate determine fit.