Written by: Team PredictiveHire
Why HR should implement predictive hiring - and how to get started
You know the common definition of insanity? The one where the same thing gets done over and over again, but the end result doesn’t change? It might not be a big deal when talking about your daily commute, but taking the same old approach to hire key personnel could be an expensive mistake.
Industry studies estimate bad hires cost up to 2.5 times the dollar amount of that person’s salary – and the damage doesn’t end there. Mismatched employees disrupt workplace chemistry, productivity, and profitability.
In response to poor hiring decisions, a growing number of companies now employ predictive screening and hiring models. Engaging predictive analytics and artificial intelligence (AI) – or algorithms that ‘think’ like humans – to help with the legwork historically performed by recruiters.
AI and predictive analytics look at historical data and then apply the learnings to new data to predict future outcomes. So, predictive hiring models can predict who will make it through the interview process, outperform their peers and still be around a few years down the road.
Progressive HR professionals have realised the potential of predictive and data-driven hiring, and hiring managers seem to agree.
“Today, HR has a seat at the table, and in order to maintain that business partnership, you need to have an analytics framework.”
Andy Kaslow, CHRO, Cerberus
A 2016 survey revealed a strong desire to drive talent acquisition through data and analytics. Two hundred executives at large U.S. firms want technology to play a bigger part in the hiring process. And the clamour for analytics isn’t confined to a younger crowd. Two-thirds of decision-makers who desire data-driven solutions fell between the ages of 45-64.
So, why isn’t everyone doing it?
Although there is a general consensus that data-driven and predictive hiring will make hiring decisions more accurate, many HR professionals still view it as cumbersome and costly to implement.
And it can be true.
Understanding the data needed to make an impact, and figuring out the best techniques and algorithms to use is difficult.
And it can be expensive to hire data scientists, and other key technical personnel needed to implement a full scale HR analytics system.
But, there’s no need to go it alone or to do it all at once.
Rather than setting up in-house HR analytics teams, most companies opt to engage a vendor who specialises in custom predictive screening and hiring models. Finding a vendor that works with you to solve your hiring challenges will significantly cut cost and time to implement.
So, if you are considering implementing predictive hiring, we have put together a few tips to ensure you get off to a great start.
1. Define what you want to improve
The crucial first step of any successful project is to define what that success looks like. And implementing predictive hiring isn’t any different.
Have a think about the biggest issue your organisation is facing at the moment that better hiring decisions will solve.
For example, you might have the issue that a lot of new hires are leaving your organisation after a few months. Or you might have a company culture in need of strengthening, and need to hire people who fit with your ideal culture.
When you have honed in on the issue you want to solve, you also need to start thinking about the data that will be required to solve your challenge.
To give you an indication of the type of data you might need, consider these examples;
- If you have an issue with turnover, the data needed would be employees’ start and end date.
- Or, if you were looking to improve your customer service level at a call centre, you would need some sort of customer service KPI data – for example NPS.
(These indications are based on the data required if you were working with us at PredictiveHire)
2. Find a shortlist of vendors
After defining the issue you want to address with predictive hiring, it is time to find a shortlist of vendors that can help you achieve your goal.
Make sure you look for vendors who are able to build predictive hiring models focused on your specific issues, whilst making sure the candidate experience isn’t compromised.
3. Perform your due diligence
When you have your shortlist of vendors narrowed down, make sure you perform your due diligence. Some vendors will be a better fit for the challenge you wish to solve with your predictive hiring model.
Make sure your shortlisted vendors address these key questions;
- What are they basing their predictions on? Predictive hiring experts should be able to tell you the scientific basis of their predictions.
- How do they address the potential of existing biases being incorporated into the algorithms?
- How do they fit in with your current hiring process? Can they fit in without causing too much disruption/change?
- What is the user experience of their product like? How about the candidate experience? For a predictive hiring solution to be useful it should be easier to use than not to use.
- How do they support you during implementation, on-boarding and roll-out of the tool?
- What is required from you in order to maximise the project outcome?
All of these questions are important to address to ensure the project’s success.
Implementing new software and processes will always require some level of change management, for example; following the ADKAR or Kotter change management approaches. Make sure you are comfortable with the level of support the vendor will offer you during the roll-out.
Following these three steps will ensure you are off to a good start with your predictive hiring project – and can start reaping the rewards quickly.
Predictive HR analytics is here to stay.
Resisting this change may put your company at a distinct disadvantage in the marketplace.
A recent MGI study found that AI can significantly improve the bottom line for businesses willing to incorporate them into their core functions. And the time really is now. Early adopters will enjoy a significant data-advantage in only a few years.
“[Leading businesses] use multiple AI technologies across multiple functions. As these firms expand AI adoption and acquire more data, laggards will find it harder to catch up.”
McKinsey Global Institute, June 2017
In the words of Gartner Research’s senior vice president Peter Sondergaard, “Information is the oil of the 21st century, and analytics is the combustion engine.”