This article was originally published on this site

I work for a small PR firm, but chatbots have always been a fascination of mine; how do they work? How are they able to generate human-like responses? Why are they so helpful on some websites and so frustrating on others? Lately, I’ve been doing some of my own research, specifically into what it would take to make my own bot and do some trial and error. To my surprise, I was able to find a bevy of helpful resources at my disposal. Amazon recently launched a bot information site that answered all my bot questions and walked me through the process, and there are countless websites out there that can help with bot hosting and building. Once I started to learn more, I started to get the bot itch – I wanted to build my own, and I wanted to test it out on my coworkers. Never one to ignore my own great ideas, I did just that.

Essentially, the bot I built was pretty basic. For a small fee, I was able to construct an extremely simple chatbot named Chaz, who would pop up in the Google Hangouts chat window of my coworkers and ask them simple questions throughout the week. Thankfully, my coworkers agreed to this little experiment, despite the slight annoyance of having an automated system hurling queries at them. The questions all related to personal experience and company culture: How are you feeling this week? What do you think of the current state of things in the office? How is your workload? The results came back to me at the end of the week, totally anonymously. With each question, the user would be able to choose from four possible responses. For example, if the question was “How are you feeling this week?” the possible responses would be something along the lines of “Great, good, below average, bad.” Not super in-depth, but at least it would serve as a decent tool to take the temperature of the overall mood of the office.

The response was overwhelming. To be totally honest, I really only expected a few people to participate. Sure, the chatbot annoyed you until you answered, but you could easily close the window and go on with your day’s work. Judging by the results, it seemed like people actually wanted to respond and vent a little about how they felt, despite the rather limited response options offered by my bot. I think the reason for this is simple – the bot represented an opportunity to unload with some distance. Let me explain a little more, as that probably doesn’t make a whole lot of sense. Imagine you’re at the water cooler (or your more modern office’s equivalent) and there’s an issue that’s been bothering you. The culture around the office has taken a dip, as people’s workloads have increased and overall mood has declined. You want to bring it up to your coworkers, but you don’t want to step on anyone’s toes or make it seem like you are the resident complainer in the office. What’s a better outlet than a human-like chatbot, which allows you to let out your frustrations in an anonymous manner?

While this small experiment was certainly interesting, I think that there are wider HR implications here. For one, there are wider implications at my company, as I plan on implementing this chatbot permanently due to the overwhelmingly positive response. It also got me thinking – if I could accomplish so much with just a simple bot I built using guides on the internet, what else could be streamlined? There are a thousand moving components that require constant maintenance for a human resources department to function effectively. What if, and I know this is going to sound sacrilegious, some of the more basic components could be automated? I’m not suggesting that we take the “human” out of “human resources,” at least not totally. I am suggesting giving employees a little bit of space to vent and complain to someone (or something, in this case) that they feel they can be totally honest with. An extremely important aspect of any HR department’s workload is hiring, which may seem a little bit difficult to automate from the outset. However, I think even this process can be streamlined with chatbots.

While this might not be as relevant to smaller companies, chatbots can be a great tool during large-scale hiring campaigns. Candidates will invariably be sending in follow-ups and questions by the dozen, swamping the already heavy workload of the average corporate HR representative. Here’s where the chatbot comes in – it can be scripted to provide the answers to frequently asked questions, taking some of the correspondence weight off of the human reps who would normally be taking care of them. On top of this, in many cases, the candidate might be nervous and, despite having an important question, wouldn’t want to email HR and bring it to their attention. This way, they can get their questions answered in a low stakes environment, and everyone comes out happy. Bots can also be put to use in the onboarding process, or to help employees (especially new hires) with basic scheduling issues and other simple inquiries. From my small experiment, I was able to conclude that people aren’t afraid of bots in the workplace. Why not put a good thing to use?

If you’re looking for a key takeaway here, it’s that you shouldn’t be afraid to experiment a little with new ideas, especially when it comes to office culture and the way that things are currently being run. Chatbots might not be the answer for you, but keep an open mind – sometimes things that might sound a little odd at first can actually turn out to be valuable solutions.

This is a guest article by John Van Duyn.