One of the longstanding issues corporate recruiters and business managers have with employee screening, is the sheer amount of time it can take to complete the research needed (criminal checks, referencing, etc.) to support or challenge a business’ hiring decision. In 20 years’ time, it is entirely possible that system automation and developments in artificial intelligence will do much of this work for us, but for the time being, we are at a crossroads where systems automation and human intervention are both necessary.
Last November, the Financial Times ran an article suggesting that in 20 years’ time many jobs could be taken over by robots. In recent weeks we have heard about one UK bank that has decided to replace a number of its investment advisors with digital robo-advisors, and several other high-street banks are reportedly gearing up to do the same.
The question is, could employee screening follow suit and become a fully automated process? At its most basic, employee screening research is made up of two core elements: database research -so credit record checks, corporate record searches and criminal checks- and the sourcing of references from educational establishments, employers and other third parties.
One of the challenges with fully automating this research is that it is reliant on the prospective hire providing correct and accurate information in the first place. This sounds simple enough but, as The Risk Advisory Group’s latest research has revealed, a growing number of CVs contain discrepancies. Indeed, as many as 70% of the 5,500 CVs we analysed contained some kind of inaccuracy- up from 63% last year.
Some of these inaccuracies are basic errors such as the candidate entering an incorrect date of birth, putting a surname in the wrong space or providing an inaccurate address. But these can lead to bigger problems if you are using an Automated Protocol Interface (API) where, for example, a credit check is triggered within seconds of a person submitting their application form. If the information going into the system isn’t correct then you can’t expect to get accurate results. This will incur both cost and time to resolve- but you could argue that employing one or two people to manage inaccuracies is a lot less labour intensive than having a team of people manually entering the data, which itself is open to human error.
The referencing aspect of the screening process is an entirely different matter. Say a potential employee provides comprehensive details of a previous employer: company name, address, contact number, line manager name, etc. A sophisticated screening system could extract this information and trigger an email (or fax, or letter) to be sent to the company, asking them for a reference. You could even configure a reference chase cycle to send out regular reminders.
However, what if the individual is deliberately trying to deceive their future employer by providing inaccurate or indeed fake contact details? There could be many scenarios where this process could be corrupted. Worst case: the individual uses a bogus firm, operating from a virtual office, with a fictitious corporate website to support a fake career history. This may sound farfetched, but it has happened.
Currently, it is true that automation can add value to the screening process and streamline some aspects of the research. However, there is nothing more enquiring than the human mind. And currently there can be no substitute for the skills of a well trained and experienced research analyst able to question and verify information provided by prospective hires.
Michael Whittington is head of employee screening at The Risk Advisory Group
