If I started this article by, let’s just say, immediately asking for your full name and home address, I can’t imagine it would go over too well.
First of all, that’s just rude.
But also, why would you ever want to give me that information? You don’t know me. You don’t trust me. There’s no obvious benefit for you. You came here to read and learn and gain value and information from me, not the other way around.
And yet…
That is precisely how most recruiting chatbots introduce themselves to candidates.
“Hi there. Can I have your first and last name and email address to better help you?”
Worse case scenario is that your aggressive introduction scared the candidate into no longer wanting to be a candidate. Best case scenario is the candidate reluctantly agreed, but you now have a relationship built on a shaky foundation of transactions and an uneven transferral of value that you will need to overcome during subsequent steps in the candidate journey. It’s like starting a marathon and then immediately stepping into a bear trap; sure, you might reach the finish line, but it’s going to be a heck of a lot harder now.
With this approach to the candidate capture process, the cart is firmly out in front of the horse. Context is everything with humans; at this point in the candidate journey, we should be providing the help first — and then giving the candidate agency to determine where the conversation goes from there. If they want to give us their personal information, that’s great. If they just want to ask a few questions and leave, that’s great, too. The important thing is that the candidates have the reins; they can ask what they want, when they want, and how they want … and the AI is smart enough to know the difference.
Context matters.
Of course, the problem was that until recently AI in general (let alone TA) wasn’t smart enough to deliver an experience like that. Including ours, to be honest. We’ve always called our AI an “assistant” to delineate between its lesser, more chatbotty brethren, but there were certainly still limitations to how much the assistant was actually able to assist. Our assistant — who we call Oliva — was friendlier and nimbler, but still had plenty of limitations.
Not anymore.
The emergence of generative AI and the widespread use of tools like ChatGPT have changed the game and raised the bar. So we raised ours, too, with an upgrade we call “contextual conversations.”
Contextual conversations leverage generative AI and allow our AI assistant to communicate with candidates much like you can with best-in-class AI tools. True to its name, it provides a level of nuance, sophistication, and context that is helping our clients deliver the best candidate experiences in the world. Gone are the days of career site chatbots engaging candidates with primitive, binary responses and limited knowledge bases. The age of context is here.
What “contextual awareness” means in the context of recruiting.
OK, let’s start over.
My name is Adam. How can I best help you today? Oh, you would like to know more about the importance of contextual awareness in recruiting? I can absolutely answer that for you:
See, at the end of the day, recruiting is about trust. And trust is built through having great conversations with people. And it’s the little things being just a little off that erode trust, while flourishes of magic can build it. A great conversationalist (or even just an average one, really) hears a question, processes it, and then not just answers it, but answers it and adds in supplemental information that is likely to be interesting and relevant to the other person, which that person can then react and reply to. And so on.
Contextual responses leverage generative AI and allow our AI assistant to communicate with candidates much like you can with best-in-class AI tools.
I know I don’t need to explain what talking is to you (you’ve probably done it once or twice), but for some reason we’ve mostly resigned ourselves to the conversations candidates have with AI tools not being this way. Most of the time they are the complete opposite. Here's a pretty standard conversation you would come to expect from a chatbot that sits on a career site, reimagined as if it was an interaction with an actual human being.
Candidate: “I’m a new mom. Do you offer work from home?”
Recruiter: “Yes.”
Candidate: “OK, and I can leave early to pick up my kids from school?”
Recruiter: “I’m sorry, I don’t understand. Ask me something else.”
Candidate: “Uh, no thanks.”
Now, if someone in real life actually responded to your questions like that, you would turn around and run in the other direction. And yet the expectation is for candidates to just … deal with it; ignore the clunky awkwardness, navigate around tons of barriers and unanswerable questions, and somehow make their way to an actual job application.
That was acceptable a few years ago (though it was never actually good). With contextual conversations, conversations now look like this:
Nimble.
Value-driven.
Accurate.
Human.
Which is exactly the kind of conversation you’d expect if your best recruiter was leading it. Now you can create that at scale, for thousands of candidates at once.
How it works, all happening in microseconds:
- First, the AI assistant detects what type of conversation we’re having — is the candidate asking a question, looking to schedule an interview, giving me information, or something else? When we determine that it’s a question, we look for context in the entire conversation to better understand the context of the question.
- Then, we go get an answer from our knowledge bases — which is a combination of specific content that our clients choose to upload and parts of knowledge found in a proprietary, self-hosted language model we built specifically for the recruiting industry — that fit the context of the conversation. With these pieces combined, the assistant generates a natural response that blends accurate information with personalized flourishes to create conversational magic.
- An important thing to note for those concerned the AI might go rogue: We’re not actually generating anything new — responses are all based on pre-approved information. There will be no complete surprises. If a candidate asks a question about PTO, the assistant will provide the information the employer wants … just in a way that feels personalized and makes sense to the context in which the question was asked. And if you want certain sensitive questions to have exact, predetermined answers, word for word, you can build those in, too.
The immense value contextual awareness can drive for your business.
We know this type of conversational approach works when it comes to talking to, screening, and scheduling candidates, because they tell us it does.
Our conversational recruiting software sees an over 96% candidate satisfaction rate. It reduces drop-off, no-shows, and time-to-hire, and it leaves candidates feeling happy. And this was before we implemented generative AI.
And now?
Well, we’re still gathering data as we enter this new frontier. But if we work backwards from what I explained earlier: Contextual awareness powers great conversations, great conversations build trust, and trust leads to candidates who want to come work for you.
The new gold standard for AI in recruiting is contextual responses; it’s not being able to sense a drop-off between what a human would say and what the AI actually says; it’s being able to understand the context of a conversation — no matter if it’s about the weather or brand values or open roles — and deliver responses that engage, delight, and continue the candidate down a positive path.
A path without bear traps. Yeah, that’s a race we’re all interested in running.
