This is a tweet from Glen Cathey quoting Shannon Pritchett, Editor of SourceCon. Were you listening? If you weren’t paying attention before you should be paying attention right now, you see, the chatbots are coming.
— 𝙶𝚕𝚎𝚗 𝙲𝚊𝚝𝚑𝚎𝚢 (@GlenCathey) September 17, 2015
Recently Google unveiled an AI that can have phone conversations with strangers. Recruitment process automation is officially here, and it has serious implications for sourcing. Now chatbots aren’t necessarily new in the recruitment process, but the level of sophistication they are starting to achieve is admittedly impressive.
We are used to AI hype in sourcing, and there is a segment of our profession that has embraced AI-powered sourcing solutions but what is coming next is an evolution in the way that companies build talent pipelines.
As a sourcer, there are two things to consider when it comes to AI and sourcing. The first is that the AI-powered sourcing tools that we are accustomed to requiring input from us as sourcers. It requires the user to have at least a basic understanding of search and how search works. The purposes of the tools are to help us find the best talent in the least amount of time.
They also give us a way to connect with that talent, but beyond providing contact information, they have not helped us directly engage with candidates. This application of AI in sourcing has mostly been complimentary. However, the nature of that relationship between sourcing an AI is about to change.
Before I get into the two things, I think you need to be aware of I think it is essential for me to define some of the areas of AI so that we are all thinking about the same thing. AI-powered sourcing tools to date have taken advantage of an aspect of AI known as machine learning. Over time these tools actually can get better and understand what “good” looks like. While mostly a “good” thing Glen Cathey recently wrote about the Risk of Bias in AI Sourcing Matching Solutions.
Chatbots also all under the AI umbrella and while they also take advantage of machine learning, they also are programmed with what is known as NLP or natural language processing. NLP is the area of AI dedicated to communication between humans and machines. NLP is designed to not look at what specific words someone is using but what they mean. NLP is designed to look for and understand the intent. Once the chatbot understands what you intend it, then it attempts to select a relevant answer that moves the conversation forward in a meaningful way. The first application of chatbot technology has been benign enough.
At first, they were deployed in place of the dreaded FAQ, and in some cases, they were able to answer basic questions and let a candidate know the status of their application. The second application of chatbot technology has been to replace or augment the application process. Instead of filling out a form a chatbot engages candidates over text and gathers the information required for the formal application. That in and of itself has been extremely effective but is not the ultimate destination of chatbot technology.
Article Continues Below
Dice's 2019 Tech Salary Report
The second thing to be aware of is that chatbots have become increasingly sophisticated. Chatbots are now being programmed to engage with passive candidates and conduct initial outreach. The current products I’ve tested do feel spammy, and they are expensive but the bottom line is that they function and are available for purchase right now.
I wrote in Sourcing is the New Recruiting that the partnership between sourcing and recruitment marketing is going to become increasingly important. From where I sit now, as a sourcer working in the field filling reqs day to day just like you, is that recruitment marketing and sourcing are going to become the same thing. Sourcers will be the ones who give the chatbots the scripts and monitor them for effectiveness and where required, jump in and do outreach.
They will have to know marketing because most of the communication from chatbots is going to be branded content. Additionally sourcers are going to be the ones who conduct the searches and tell the chatbot who to engage. Instead of sending template emails sourcers are most likely the people to identify talent and send the chatbot out to engage the prospect. The responses to outreach are going to be programmed by sourcers and monitored by marketers.
That said there will always be a demand for human sourcers, but the expectation will be that they are expert communicators and a response rate of 70% or more will be the reasonable expectation of a sourcer. I do not think sourcing is going anywhere for a long time, but I do believe that the tools we use for our work will. The reason to send a chatbot out to engage with candidates is apparent. A human sourcer can only engage so many people at a time. A chatbot can reach many people with a semi-customized message nearly instantly. A chatbot is also available on Sunday at 2PM where a human sourcer should be doing something that isn’t sourcing related.
I think the best advice I can pass to you remains the advice I received some time ago from Ninh Tran COO of Hiretual, “(in the future) It is going to be humans partnering with technology to make things simpler and faster. If you don’t embrace it, that is fine, but if there is a person that wants to keep their competitive advantage for years to come, becoming a new technology adaptor is the way to go.”