There is nowhere to hide. Everything we do, write, and say is available to those who know how to access it. Software, cookies and algorithms are monitoring and tracking what we purchase. They are watching our social media posts and comments, and noting what we like and do not like on Facebook, who we are connected to on LinkedIn, which videos we like, what music we listen to, and where we go. Search capabilities have expanded along with growth in computer capability, speed, and storage.
Searching for candidates using Boolean or by scraping job boards may be old hat soon. Predictive analytics and cognitive computing are replacing this type of search and also helping in assessing and engaging candidates.
A variety of tools are tracking us even if we do not use social media as long we access the Internet, use any e-commerce applications, a smartphone, or subscribe to a cable network. And as more and more apps spread into areas of home security and personal health an even more complete picture of us emerges. Apps ranging from Instagram and Pinterest to pedometers and travel apps such as Tripit are gathering prodigious amounts of information, which when combined with some intelligence, predict or know what you will buy, what your personality is like, where you travel to, how much you exercise and many other attributes.
[bctt tweet=”Predictive & tracking technologies replacing traditional sourcing methods – should we be concerned?” username=”ATCevent”]
Recruiting applications are also tapping into this stream of data to learn more about you and to create an assessment of your skills, personality, experience, and general knowledge. These apps will soon be able to predict what work you are best suited for and which jobs or positions match best your skills and interests. And they will show you specific jobs when the match level reaches a predetermined level of probability that you will be interested and accept an offer. The algorithms or formulas that determine this interest will use the science of cognitive computing to refine themselves, learning from the actions you and others take.
Candidate Sourcing & Searching
Finding qualified candidates has been a challenge for recruiters for at least the past decade. Job requirements have increased, the existing workforce is aging, and the pool of skilled applicants has narrowed, creating a perceived shortage of candidates (see my recent article on this). The shortage is partiality caused by our inability to access the total pool of potential candidates because of lack of knowledge about where they are or what their skills are. Currently, the most successful sourcers use advanced Boolean search techniques in an attempt to find people who are not actively looking for a job but who have the desired skills.
New software tools monitor websites, social media streams, seek out papers and books written recently, track speakers at events and conferences, and look for mentions of people in emails and other communications. By their ability to scan huge amounts of data and extract information about the skills and even personality of potential candidates they can then match them up to whatever jobs are posted.
In addition, these tools communicate with a potential candidate by presenting them with personalised messages and by showing them opportunities in ways that are highly likely to interest them. For some people, the messages might be video, if that is what is most likely to interest them. For others, it might be a text or a link. Each outreach will be tailored to get the highest response possible. The number of candidate relationship management (CRM) software firms has expanded over the past few years, as has their capabilities. They provide meaningful, tailored messaging with minimal human input.
Chatbots are also appearing on career sites. These are software applications in the form of popup boxes where a “person” offers to answer questions about services or products. These are automated and the “person” is a computer that can provide customers with immediate answers to questions. This same ability is now available for candidates. By using semantic analysis, the computer understands what is asked, searches databases and previous responses to answer intelligently. Because these bots are usually part of a cognitive computer system they learn from the actions people take and get better overtime. And usually the user has no idea that they are speaking with a computer.
The U.S. Army has been using a chatbot for recruitment since 2006, called SGT STAR. According to the army, this chatbot has done “the work of 55 human recruiters. Over the last five years, he has answered 10.5 million questions in 2.8 million chat sessions”.
The data SGT STAR has gathered is invaluable in making decisions on which recruiting messages and techniques have been the most successful in turning the visitors into candidates for enlistment.
Other tools use data gathered internally from human resource databases to make predictions about which employees are most likely to leave. These tools look at data such as when promotions were handed out, economic factors, the length of tenure, changes in education, industry trends and other data to make its predictions. The software can recommend a series of interventions from increasing salary to just having a talk when the employees are identified as likely to leave. The software can be predictive and assign probabilities to how effective each intervention will be.
[bctt tweet=”We need a code of ethics on the use of predictive & tracking technologies says @kevinwheeler” username=”ATCevent”]
Many tools have emerged over the past few years that offer insight into a candidate’s personality by scraping data from social profiles or from written text the candidate has created and posted on the web. IBM has developed a tool to provide a personality assessment that measures personality using the Big Five personality characteristics as well as values and needs. And the tool is learning and updating the algorithms it uses all the time.
HireVue, a video interviewing platform, offers a product called Insights that parses data in an interview, analyses it for language use, sentiment, and expression, as well as for personality and engagement to compare candidates with the top performers in the organisation. Another product called Crystalknows, analyses an email as you write it and provides advice on how to make your message more effective based on analysing the person’s social media profile.
Many other firms are offering personality-type analysis of candidates based on Facebook, Twitter, blogs and other data that exist about the candidate.
All of this happens in the background and usually without the candidate being aware that any tracking or analysis is happening. While this is providing recruiters with more candidates and candidates with better service by more closely matching their backgrounds and interests to opportunities, the software also brings potential problems of invasion of privacy, lack of transparency, and a bit of the fear of “big brother” watching us all the time. How this will play out remains to be seen, but there will most likely be a backlash soon as knowledge of what is happening becomes known.
Meanwhile, the recruiting world is and will be using all of these capabilities and we should begin developing a code of ethics about how we use them and what level of transparency is best.
Leave a Reply