From ELIZA the 1964 ‘chatterbot’ that created an illusion of human understanding to today’s advanced AI conversational assistants, chatbots have come a long way. How do you find one that’s perfect for your business?
Short for Artificial Linguistic Internet Computer Entity, ALICE was the world’s earliest chatbot, created in 1995. While ALICE has impressive language processing and linguistic prowess, it cannot be used for much more than having a fun conversation with a robot.
Today’s chatbots are far more powerful than ALICE. Unlike early rule-based bots that were programmed to follow a script, today’s chatbots are more like conversational assistants. They use Natural Language Understanding (NLU) and Natural Language Processing (NLP) to understand the user’s intent, engage them in a conversation, and either guide them to a solution or collect useful information for your business.
Recruiting chatbots are an excellent example of these. They work as digital recruiters available 24x7 to:
respond to queries from job seekers, whatever time or day these come in.
screen candidate profiles (for instance, do a check of their location, pay or qualifications.)
reach out to your talent database when a new role opens up.
Perhaps you are considering taking a recruiting chatbot out for a spin—the question is, which one?
Bots, bots everywhere
There are quite a few chatbots available today but not all are equal. In this nifty guide, we want to help you evaluate your options and arrive at the right choice for your needs.
“A chatbot is something everyone needs to evaluate. Especially considering the potential it has to impact your candidate experience and your recruiters’ day-to-day activities. As a product, it has the ability to transform the industry.” Greg Steering, Customer Experience Supervisor at PrideStaff
Start with a goal-to-feature mapping exercise. What prompted you to look for a recruiting chatbot? Perhaps your recruiters are spending nearly two-thirds of their time answering repetitive questions. Perhaps you are missing out on candidate engagement opportunities because of timing or delays. Or perhaps as a user, you have seen what chatbots can do and want to leverage that tech for your business.
Starting from the first stage of your hiring funnel, put down how chatbots can make things more efficient at every step. Then look internally at the way your hiring team is currently working. What gaps or areas of improvement do you see? List out how a chabot might be able to help. Lastly, go back to the business goals you set for the year. Could a chatbot help you achieve these goals faster?
Now put all of these together—this is your product wishlist. When you are evaluating chatbots, compare their listed capabilities with this to see if it is for you. Here are 8 key questions to ask during this process.
#1 What kind of candidate experience can the chatbot deliver?
In recruiting, candidate experience is everything. The last thing you want is for a bot to frustrate a job seeker who will then bounce away with a negative experience. So, evaluate if conversations with the bot feel natural.
Is it able to answer typical candidate queries effectively without going into loops? When the bot is unable to find an answer, what response does it give? (Alexa says “I am having trouble understanding you.”) And is it able to suggest an alternative to the candidate? (“I’m sorry. I didn’t quite get that. Is your question related to one of these? (a) Pay (b) Location (c) Job role.”) In essence, ensure that candidates have easy, effective conversations with the bot.
#2 Does the chatbot use NLP?
Rule-based bots are designed to answer the most frequently asked questions. They give standardized responses to these and for anything outside their decision flow, ask the user to try again. But recruiting chatbots need to be much more powerful because this is an area where the human or personal touch is very important.
We recommend investing in an AI chatbot that uses NLP. NLP is a confluence of AI and linguistics that helps machines understand, interpret, and use human language. Which is why NLP-powered chatbots sound much more like actual humans.
They don’t just recognize words and phrases used by the candidate but actually understand the context and intent of the question. They are able to answer more complex questions using natural-language responses and the more you ‘train’ the bot, the better it gets at this. A
powerful recruiting chatbot can also recognize the candidate’s mood and emotions by evaluating certain words and phrases, use of modifiers, exclamations or expletives, and more.
Texting lingo is very different from colloquial speech. But as humans, we automatically translate texts to speech and respond appropriately. So this has become the baseline expectation from chatbots as well. For example, a candidate might answer “Yup” or “Yeah” to a question or use a phrase like “That’s cool.” Can the chatbot recognize that all of these are affirmative responses to the question posed? Evaluate also for understanding of common acronyms like BRB, OK or TY.
Remember: the more natural the conversation is, the better the candidate experience.
Source: Microsoft Cloud Blogs
#3 Does the bot have reasonable decision making abilities?
When job seekers ping you, it is because they need clarification or guidance. And to be truly helpful, the bot needs to have reasonable decision making capabilities to direct the candidate the right way.
No, this doesn’t mean the bot needs to make decisions on its own. A truly effective recruiting chatbot is one that you can program with custom flows and decision trees that will guide your candidates to helpful outcomes.
#4 How does the bot score on customization & integrability?
Let’s say you really like the chatbot. But how easy is it to implement? Any chatbot you choose needs to seamlessly integrate with your existing ATS and other recruiting tech such as email, texting, and marketing platforms. It’s only when all of these come together that you will be able to get tremendous ROI.
You should also be able to customize the chatbot to suit your brand and business. As Mithun Gunalan, product lead for Sense Chatbot, says, “At the heart of your conversational assistant is, well, you! Flexibility is of essence here. Putting “you” into your chatbot means you need control over your conversational builder.”
#5 How scalable is the chatbot?
While chatbots can show results soon after implementation, they are definitely a long term investment. You might start with a pilot, testing the bot in one area at first, but scale up soon for other parts of your business. Your goals could also change from this year to the next, resulting in a whole other set of priorities. The chatbot you choose must, thus, grow with your business.
A lot of our clients start out using Sense Chatbot to answer questions from job seekers on their platform. (“How much does this role pay?” “Can I apply to work part-time?” “Are X remote-friendly”) Once they start seeing the time savings and opportunity areas for recruiters, we help them deploy Sense Chatbot in other areas. For example, re-activating their existing database. One staffing client used Sense to ping inactive candidates in their ATS to ask if they were currently open to new opportunities and if so, gather additional information such as location and expected pay. This helped them identify a high potential segment for their recruiters to follow up.
So the next check we recommend is this: how well can the bot support your future needs? How scalable, flexible, and adaptable is it to changing business requirements? Rank the bots you are evaluating based on this to come up with a shortlist.
#6 Is the chatbot a good learner?
Yes, bots have come a long way since the days of ELIZA and ALICE—but they are still learning. In fact, they are designed to continually learn and up their capabilities. When you are evaluating chatbots, ask if it is constantly evolving? If it is, you will get much more value out of it the more you use it.
The Sense Chatbot, to take an example, leverages Natural Language Processing (NLP), Natural Language Understanding (NLU), artificial intelligence (AI) and learns from every interaction and today, understands natural language, acronyms, industry lingo, and more. This makes the conversations with the bot far more natural and seamless for candidates. Some of our clients have seen NPS scores of 4.6 on 5 for the Sense Chatbot, indicating that candidates were thoroughly happy with their bot chat!
#7 Is it diversity-friendly?
Organizations everywhere are recognizing the importance of diversity and inclusion and putting into practice programs to make this happen. Eliminating bias right from the hiring stage is the key to a diverse, inclusive workplace. Unfortunately, just as humans come with personal biases and prejudices, bots can mess up too if not designed with inclusion in mind.
Remember Tay, the Twitter bot tweeting about Hitler and 9/11 back in 2016 and enraging the public? Programmed right, recruiting chatbots can actually champion your diversity cause by being ‘blind’ to personal identifying information (gender, age, race, sexual orientation) during early screening and evaluate candidates on their merit and suitability for the role.
#8 Who are the chatbot makers?
Once you have a shortlist of the chatbots you really like, it is time to evaluate the makers. They are the folks you will work with to plan and implement your AI-powered communication strategy. You need them to be great at what they do, professional, trustworthy, and always ready to pitch in if there’s a glitch.
We believe that what matters the most here is industry expertise. Because only staffing industry insiders really know the practical challenges faced by recruiters. A chatbot trained for customer support, for instance, will not help recruiting teams. Check how long the makers have been in the space, who’s using their products, and their track record.
Once your conversation begins, ask for case studies. Speak with previous clients. These will help you understand exactly what it takes to implement a recruiting chatbot and what benefits it brings.
Lastly, ask about the data security measures put in place to ensure that the chatbot is not vulnerable to breaches. At Sense, for example, we use end-to-end encryption, identity authentication, and secure protocols to ensure that your data is always protected.
Created by staffing industry veterans for recruiters, Sense Chatbot has helped dozens of businesses improve recruiting efficiency and speed up the hiring process. From enriching the data in your talent base to screening, interview scheduling, and reactivating dormant talent, it can help you do so much. For a walkthrough, schedule a demo today.