Companies have a wealth of information about their employees. Most HR systems can easily capture performance data, retention data and employee engagement. This data lends itself well to using predictive analytics to hire people using statistical models instead of pure intuition.
Collecting and organising data on job performance, job tenure and engagement of existing employees can predict various things e.g. who’s likely to leave the company, expect a promotion or ask for a pay rise. In the recruitment domain, data can be used to identify which candidates are most likely to be hired, to perform well or likely to stay in the company if hired.
If you’re not already using predictive hiring analytics in your recruitment process then you’re missing a big opportunity to optimise your hiring process.
What is predictive hiring and why does it matter?
In short, predictive hiring is the use of predictive analytics in the recruitment process. Predictive analytics involves analysing large quantities of data, usually from current employees, to make predictions about which candidates will make good hires. It’s the kind of thing we do every day when making hiring decisions. The only difference is that software can do this more effectively and with a much larger quantity of data than the human brain.
As this diagram demonstrates; analytics engines take large quantities of data; sort it, and predict outcomes in the future. The more data you input, the more effective it will be.
Why use predictive hiring analytics?
The promise of predictive hiring analytics is compelling. Automating the top of the hiring funnel to identify the most suitable candidates can save companies an extraordinary amount of time. Candidate screening software like ThriveMap can save time by assessing applicants on job-relevant questions and tasks rather than sifting CVs.
Most importantly, predictive hiring is proven to be cheaper, quicker and provide better quality hires. Especially when compared with the old form of “instinctive hiring”, or going with your gut.
Furthermore, using predictive hiring analytics can reduce bias and increase the diversity of new hires. Models should only use quantifiable data such as skills and experience instead of non-job relevant factors like ethnicity, gender and background.
However, we need to proceed with caution as employee data is rarely bias-free. When using machine learning or AI to predict suitable hires it often inherits the bias of previous human-made decisions. This can systemise bias into the statistical model making it a potentially dangerous tool.
How to use predictive hiring analytics in recruitment
Predictive hiring analytics can be used in almost all areas of recruiting, from candidate sourcing to onboarding. Detailed here are three of the best ways to implement predictive hiring in your recruitment funnel.
Tracking the source channels of your best performing hires enables you to optimise your candidate attraction strategy. The time and money saved from narrowing your search away from less lucrative and therefore more costly areas can be redistributed to improve your appeal to those ideal candidates.
The self-learning elements of predictive analytics will reduce your time to hire by more efficiently isolating the factors that produce productive employees in a given role. In the graphic above from Ideal companies who use assessments to screen candidates can save 23 hours per vacancy. At ThriveMap, each of our customers each gets their own predictive hiring model. This shows you which skills, behaviours and attributes are leading to successful hires in your company. This is then used to optimise the performance of your assessment over time.
Despite the risks associated with systemic hiring bias, you can choose to feed a predictive model with data on existing employees. Performance reviews, employee surveys, attrition data, can all be fed into your hiring analytics to help predict the future job performance of your candidates. Rather than having technology making hiring decisions for you, it should be used alongside human decision-making to augment the recruitment process.
An example of predictive hiring analytics
Safelite is a US auto glass repair company. They wanted to improve the performance and retention of new hires. Following a job shadow and assessment workshop we developed an ideal candidate profile which captured the attributes required to thrive in the job. Based on that data, ThrivMap then developed a work simulation assessment to give candidates a virtual “day in the life experience” of the job. The assessment was implemented and as a result, new hire retention improved by 27% and performance by 36%.
Advances in technology means that there’s no reason why you can’t cheaply and efficiently use predictive analytics to screen, source and get a more comprehensive picture of your candidates. In this article we’ve shown you the basics of how you can use data to “predict the future” of your potential employees. This will save you from making bad hires and allowing you to invest your recruitment resources more efficiently. Why go with your gut, when you can go with the data?