It’s time to put data at the heart of the recruitment process

Phil Brown explains how ditching CVs and taking a more scientific approach will help employers avoid costly bad hires 

Of all the processes HR professionals must juggle, recruitment remains the one with the most immediate business impact. Sadly, it is also the area where HR and talent acquisition – and the business as a whole – are most likely to be unsuccessful, as almost 50 per cent of new hires are likely to fail within 18 months.

There are many points of possible failure, from how the job ad is crafted, right the way through to onboarding and training. However, when a recruitment exercise starts with screening CVs, false assumptions, hunches and bias will inevitably shape results for the worse.

The problem is that CVs do not always confine themselves to fact. A CareerBuilder study found that 46 per cent of candidates exaggerated their ability or the extent of their experience. In another survey by First Advantage, 37 per cent of CVs submitted to technology companies had inaccuracies – a full 10 per cent higher than the national average. Worse still, almost 27 per cent of those inaccuracies were considered so serious that they were a major cause for concern.

Even when intentional dishonesty is not in play, there is an inbuilt conflict. Applicants want to place themselves in the best possible light to get an interview. They may genuinely believe their own hype and impact on a company’s success, even if there is little basis in fact. So, down it goes on the CV. 

At the same time, HR is looking to deselect unsuitable candidates using a set of criteria that relies on accurate and complete data. You can see the problem. Some roles may not be well understood by those conducting the screening. There are many nuances in all but the most basic of positions, particularly for technical jobs, that only a hiring manager is likely to understand in the context of an application. What can be done?

Data is key. Look to your existing hiring KPIs to understand if they truly support ‘quality of hire’ metrics. If you don't already have such metrics in place, create them now as they will serve as a significant competitive advantage – according to LinkedIn's Global Recruiting Trends survey, 67 per cent of businesses aren't doing a good job of understanding quality of hire. Take a more scientific approach to recruitment process design. Dispense with the CV. Look for areas where data can take the place of ‘instinct’ or ‘gut’ decisions.

You should quantify your failure rate, but most importantly understand what a bad hire is costing you. According to the Recruitment & Employment Confederation, a poor hire at mid-manager level with a salary of £42,000 can cost your business upwards of £132,000. For specialist roles, expect the cost to increase by an order of magnitude.

As a first stage, invite applications via an unbiased, evidence-based assessment constructed using the specific skills and abilities required for the role. Be careful to make a distinction between those skills that are essential and those that are merely nice to have. 

Invest in an online assessment platform to automate the distribution and processing of tests. Consider adopting a blind recruitment process – most good assessment platforms should allow you to anonymise candidate data so that bias isn't reintroduced when selecting the most promising candidates to fast-track to interview.

Regularly quantify the existing skills landscape in your business. Not only will this add further rigour to your training initiatives, it will set useful benchmarks for recruitment and may also be useful for succession planning.

Phil Brown is co-founder and chief product officer at Technically Compatible