Data Driven Recruitment


The hiring of successful people is still one of the most important factors for any company, no matter how small or large your company is. Most companies look at a candidate’s resume, perform aptitude tests, interview them, and much more, to make sure that they will make the correct hiring decision.

Yet despite all these efforts, recruitment is still a messy process. An average job receives 250 applicants, yet the candidate chosen by the company fails 30%-50% of the time. Resume reviews often lead to biases, resulting in women and minorities being disadvantaged.

Moreover, the process is often frustrating for both the applicant as well as the company. The company has to wade through all these applications, review them diligently, and manage the entire process from start to finish. On the other side, applicants often review the recruitment process as poor, with 45% of applicants not even hearing back from the company.

The recent shift into data-driven recruitment is a natural shift to improve the entire recruitment process. However, there are still plenty of companies who are still a little hesitant to incorporate data into their recruitment process. These companies have several reservations about this change. They think data-driven recruitment is expensive, time-consuming, or that it takes the human aspect out of the recruitment process. These are persevering myths about data-driven recruitment that can hold a company back.

In this article, I will be going over 7 of these myths and explain why these myths are inaccurate.

1. Data-driven recruitment is expensive

Collecting and tracking recruitment data not only sounds daunting, but it also sounds expensive. However, the cost associated with data-driven recruitment is, in reality, one of its best parts. It not only helps you improve your hiring process, but it will also help you to reduce your hiring costs.

An example I heard not too long ago, was of a company where management was complaining that it took a long time for HR to hire new candidates. To analyze this, one of the recruiters created an overview with all the steps in the hiring process in Excel with the number of days it took to complete each of them. The result? Three steps severely delayed the hiring process:

  • Getting input for the function description from the manager
  • Getting the ‘go’ for the CV selection from the manager
  • Planning a date for the interview… also with the manager

By showing that the biggest bottlenecks were in the steps where the manager was involved, the recruiter was able to show that the recruitment process was delayed because it wasn’t given sufficient priority by the manager. This solution wasn’t expensive at all.

In addition, smart selection tools will allow you to decrease time to hire and lower associated costs.

2. Data won’t allow you to assess candidates’ true fit

Assessing whether a candidate is the right fit for the company is a difficult task. A well-prepared interview, combined with an experienced recruiter, used to be enough. However, the addition of assessment tools, such as IQ- and personality tests, may put a wrench in the recruitment process.

These assessment tools sometimes go against the gut instinct of a recruiter. Therefore, recruiters often thought that this type of data doesn’t allow for the true assessment of a candidate’s fit for the company. Assessment tools should be used as an afterthought, and the recruiter’s gut instinct should lead the way.

However, unfortunately for recruiters, this gut instinct has been shown to be an unreliable predictor of success for new hires. Recruiters are often biased and frequently make mistakes. Assessment tools allow for a much more unbiased look at the fit of a candidate. Furthermore, assessment tools give you the capability to test a wide range of variables that are important for the success of the new hire, such as how detail oriented they are.

Finally, these assessment tools are created by people and are therefore flawed to a certain extent. We choose variables that we think will help predict the success and performance of new hires. Recently, there has been a surge in machine learning for fit, that will allow you to to create models you never even thought of. Vendors are now successfully applying algorithms to create tests and simulations to make hiring more effective.

An example of this is the fast-growing company Pymetrics. This company developed a series of cognitive and neurological tests that are fun and easy to take and that have a direct statistical correlation to the candidate’s success in multiple roles. They can assess a plethora of traits through these tests and virtually eliminate any bias in the recruitment process. Companies like Unilever and Tesla swear by the system and it’s forecasted to take the recruitment world by storm in the next few years.

There is a risk in these machine learning approaches, as the recent controversy regarding a hiring bias at Amazon showed. Your algorithm is as good as your data – so your data should be accurate as well. This leads us to our next point.

3. Data removes the human-aspect from recruiting

This is one of the most prolific myths about data-driven recruitment. Most people think of cold, hard facts when they hear the word “data”. So, it’s no wonder recruiters often think that data-driven recruitment removes the human aspect from recruitment.

However, data should never be looked at in isolation. When you look at finance, marketing, production; All of these departments are driven by data to help inform people’s decisions. Where to prioritize, which people to promote, where to invest. These departments combine data with the expertise from people to make data-driven policies and decisions.

Likewise, data is a tool to help your recruitment function succeed in making the best hiring process. Data will allow you to see which problems arise in your recruitment process, but it won’t tell you which problems to tackle. It is up to the recruiters, the people, to make the final decisions.

4. My team has trouble understanding the data and shifting their mindset.

This myth is one of the most troubling myths for talent acquisition leaders. Even though you might want to change their recruitment process to become more data-driven, how do you get your recruitment team to understand and support this process?

GEM Recruiting AI

There are two ways you can tackle this issue. First, to make sure your team understands how to work with data, put them on a small course that teaches them how data works and the way they should approach data. This can be done at your local university, or by looking online at several courses that teach students how to understand and process data correctly.

This doesn’t have to be a course that takes months to complete and many hours of work. Often times, courses that show the basics and take 1-2 days are enough to get a recruitment team up to speed. After, they can expand upon this foundation by simply implementing data-driven recruitment processes in their work. 

Secondly, you can create support for your data-driven recruitment function by showing your team the benefits that it gives. Don’t give them a long-winded vision of how data will transform their jobs for the better down the road. Instead, look at short-term wins that immediately impact their work on a day to day basis. Do they have to give weekly reports on each vacancy? Create simple and powerful dashboards that automate this work for them, allowing them to focus on the thing that actually matters: Finding the right people for your company.

5. Only large companies benefit from data-driven recruitment

While it is true that large companies can benefit a lot from data-driven recruitment due to their size and hiring capabilities, this does not mean that small companies cannot benefit from data-driven recruitment.

Within small companies, every hiring decision matters. I’ve personally witnessed multiple thriving companies who went under due to the intake of bad hires, who brought the rest of the team down and had enormous costs associated with them. Therefore, it may be even more beneficial for small companies to adopt a more data-driven way of recruiting, to enable you to better make hiring decisions.

Smaller companies can also benefit a lot from data-driven recruitment policies. Whilst some advanced techniques like extensive candidate profiling based on existing employee profiles are impossible, they can still leverage metrics such as time to hire and cost per hire.

Furthermore, small companies looking to save their budget can particularly benefit from low-cost recruiting tactics. Recruitment tools that can streamline the hiring process will allow you to save precious time from managers and employees involved in the hiring process. This blends into the next point.

6. It’s too difficult to get the recruitment data needed

While it may seem daunting to start with the collection of recruitment data, it’s easier than people often realize. The main thing you need to do is to select several important recruitment metrics. Recruitment metrics are measurements to track hiring success and optimize the hiring process.

While you could track dozens of different recruitment metrics, it is best to keep your selection small. This will allow you to focus on the ones that will truly help your hiring process, making it easier for you to get the recruitment data needed and create an impact.

Metrics that are both easy and important to measure are:

  • Time to hire
  • Cost per hire
  • Sourcing channel effectiveness
  • Quality of hire

Smart job promotion techniques through highly targetted LinkedIn advertisement tools can also be very effective at bringing in suitable candidates at low cost. In my experience, LinkedIn can be a very effective channel for highly educated professionals. Depending on your candidate profile, other mediums could also be effective. For example, we’ve had great success in finding suitable interns in specific study and university Facebook groups that we wouldn’t have been able to find otherwise. Posting in such groups is free.

Some of this data might already be present in your applicant tracking system. Taking a look at the data these tools collect can already provide you with a lot of input.

7. It takes too long to analyze all of the recruitment data

While it does take time to analyze your recruitment data, this does not mean that you shouldn’t do it. Data analyses is a skill, and through practice, you will be able to become more efficient at it. In addition, it can be very simple, just like I showed with the example before.

If you spend all your time measuring recruitment metrics and not enough time actually recruiting, you’re doing something wrong. It’s all about finding the right balance between analyzing your recruitment data and putting your insights gathered from your analyses to work.

Furthermore, the time you spend now on creating a well-thought-out data-structure will help you in the long run. If you change the processes and tools you use now, you’ll be ahead of the game in a year. The time you will save then will more than make up for the investment you will do now.

Finally, more and more tools are coming out that help us understand the data that we’ve gathered. Think of tools such as Hirevue, BreezyHR, Workable, and more. These tools allow you to quickly analyze your data, pinpointing exactly where you need to look and improve your efficiency.

While it may seem daunting to start with data-driven recruitment, this should not stop you from pursuing it. While it may be easy to dismiss it with the myths mentioned in this article, data-driven recruitment holds immense potential. It should be an assistant to your own experience, help you reduce your hiring costs and improve your recruitment process.





Erik van Vulpen is the founder of Analytics in HR (AIHR). He is writer, speaker, and trainer on people analytics. Erik is an instructor for the HR Analytics Academy and has extensive experience in the application of HR analytics.