Let me let you in on a not-so-hidden secret. If you’re one of the many teams out there not using analytics in recruitment, then you’re missing out on a virtual gold mine of valuable information. I don’t mean to sound dramatic, but a data-driven approach to recruitment is an absolute must in today’s talent environment. Especially if you hope to consistently and efficiently hire the best candidates. And in this post, we’re going to show you how to get started.
If you’re a bit hesitant about jumping into the scary world of recruitment analytics, then you may be interested in hearing some of the companies who are actively embracing these tactics. Names like Google, Cisco, Sprint, and Deloitte all use recruitment analytics to drive their decision making and hiring processes, and do so with industry-leading success.
That’s because the benefits of analytics in recruitment over traditional hiring are endless. Here are just some:
- It provides objective visibility into the effectiveness and value of your recruitment activities.
- It helps keep track of high-potential candidates, allowing you to actively nurture possible future hires.
- It lets you create a robust talent pool, or a permanent record of all candidates or hires that you can consistently come back to.
- It unlocks the potential to learn from and improve processes.
- It enables proactive recruitment (rather than reactive recruitment) to drive better and more timely hiring decisions.
- It lets you predict which candidates will be high performers, and which ones may be bad hires.
Before we move into how to take advantage of these benefits, let’s first explain what we mean by recruitment analytics.
What is recruitment analytics?
Recruitment analytics enables organizations to hire faster using a combination of data and predictive analysis. Analytics and the techniques behind it are as broad as the technologies available to the recruitment industry.
For the sake of this article, we’ll focus on predictive analytics in recruitment. And to get you excited to learn more, you may be interested to know that predictive analytics has been shown to save up to 23 hours per week in manual labour, mostly consisting of shortlisting and pre-screening candidates.
Do I have your attention? Good! Let’s get started.
What is predictive analytics in recruitment?
Simply put, predictive analytics in recruitment is the process of using historical data to make predictions about future hiring activities and candidates. It’s all about collecting and analyzing data using statistics, machine learning, and modeling techniques to best predict what could happen under specific scenarios.
This is made possible by the collection of large volumes of data from a variety of platforms. As well as thanks to the increasing use of workforce analytics solutions to measure variables like performance and engagement. The presence of these large data pools means that complex predictive algorithms can be applied to forecast future results.
Take an Applicant Tracking System (ATS), for example. An ATS like Recruitee is able to (with your help) gather various external inputs about a given candidate from sources like their CV, cover letter, assessments and pre-screens, social media, and so on.
Then, using predictive modelling – the basics of which I’ll outline shortly – you can analyze these inputs to determine if that candidate will be a good or bad hire.
This same principle can be applied to broader recruitment activities as well, resulting in:
- Hiring process quality enhancements
- Intelligent and efficient sourcing
- Faster and more targeted hiring
This differs from traditional hiring techniques in one fundamental way. It allows for almost absolute objectivity in your decision-making, rather than a more traditional “going with the gut” approach. By basing hiring decisions primarily on data and algorithm-based predictions, recruiters can remove their own biases from the process, which in turn can yield more consistent and better results.
What can you measure and improve using predictive analytics?
The data that you can measure using predictive analytics in recruitment is a broad as your company’s unique inputs and sources of information. Essentially, any variable associated with a candidate or process can be collected, analyzed and measured.
To give you an idea of what these possibilities look like, here are some of the categories and questions that predictive analytics can provide provide insights into:
- Candidate sources: which are your most effective sourcing platforms (job boards, social media, referrals, and so on).
- Candidate screening: how long the process takes, which candidate screening techniques are effective, and which aren’t.
- Lead times: how long it takes to go from application to offer and what effect that has on drop off rates.
- Future employment needs: what positions are likely to be needed or become vacant in the future and what will the hiring manager’s needs be.
- Future employee performance: how likely a candidate or new hire is to perform well on the job.
- Retention rates: how long a new hire will stay will the company or how likely it is that other candidates will leave.
- Hiring bottlenecks: where roadblocks in the hiring process regularly occur, what their impact is, and how to fix them.
- Urgency of hiring: which roles and skills are needed most urgently to meet company needs.
Predictive analytics can answer these questions by leveraging complex technologies and your data inputs to find trends and indicators of future behaviour and results. It should be noted that predictive analytics is only as effective as the data you provide to your tech stack, and how well you measure and respond to the data that it provides.
How to get started with predictive analytics in recruitment
Now that you’re aware the what’s possible with predictive analytics, and the benefits that it can provide, let’s talk about how to deploy these techniques in your own company.
Step 1: Choose your tech stack
There are endless HR platforms and data analytics tools on the market to choose from. The bottom line is that no one solution will suit everyone, so you’ll want to do a bit of homework to determine which is best for you.
Ideally, your tech stack for predictive analytics should be tied into your ATS or HRIS. Or, where you collect and store candidate information. Choosing an ATS that contains predictive analytics capabilities is highly convenient in keeping all raw and output data in one place.
The Recruitee ATS, for example, allows you to store all inbound candidate information on the platform, and leverage robust analytics, tracking and reporting to create actionable insights from your data. Some of Recruitee’s core capabilities for this include tracking job success, and candidate conversions, while also providing easy-to-access dashboards to track KPIs and team performance.
Whatever platform you choose, be sure that it allows you to track all relevant KPIs easily and efficiently.
Which brings us to Step 2.
Step 2: Choose your KPIs
The next step is to work with your team to determine what you want to improve, and which recruitment metrics are most important to achieving that goal. As mentioned, analytics platforms can and will collect as much or as little data as you tell them to. But, it helps to have an idea of what you’re trying to improve so that you can focus on measuring only the most relevant KPIs.
A good way to get started with this is to create a recruitment matrix. These are simple spreadsheets that break down your priority areas of improvement, and the primary metrics that relate to each. You can then weight the importance of specific metrics versus others to specify high, medium and low priority KPIs. A recruitment matrix is a great way to create a visual representation of your priorities with your team, which can then be used to upload your custom parameters into your ATS.
Examples of key areas of improvement, and specific KPIs associated with each might look like this:
- Lead Time: amount of time to work through pre-screen; amount of time from interview to hire; total amount of time from application to hire
- Candidate Sources: cost per source; hires per source; retention rate after one year per source; performance and engagement scores per source
- Quality of Hire: performance scores after one year; engagement scores after one year; competencies growth after two years, promotions after three years
As you can see, from simply brainstorming which core areas you’d like to improve, you can come up with a shortlist of KPIs that are likely to be relevant. To help with this, most analytics platforms also provide pre-determined metrics that you can use to expand on your KPI list.
At this stage, you’ll likely have a shortlist of KPIs that you think will help predict future results. From here, the next step is to apply the predictive analytics model to your data, and measure the results.
Step 3: The predictive analytics lifecycle
Once you have your tech infrastructure in place, and have worked out a plan for what to measure and why, the next step is initiative the predictive analytics lifecycle.
This cycle looks something like this:
- Collecting data
- Pre-processing (cleaning) the data
- Establishing an analysis type
- Training the model
- Performing predictions
- Acting on insights
If you’re thinking to yourself “wait, I didn’t sign up for establishing analysis types or training prediction models,” then you’re in luck. Many predictive analytics platforms in recruitment will handle these heavy-lifting phases for you.
The important parts for you to focus on are collecting the data, making sure it’s clean and accurate, and them acting on the predictions and insights that come from your platforms. This comes from ensuring that your input data only contains information that is accurate and relevant to your candidates and processes, and specifying which KPIs you’re looking to improve. By doing so, your tech stacks will be able to accurately handle the grunt work of processing the data, finding trends, and reporting on possible improvements or deficiencies.
From there, it’s a matter of continuously monitoring your data inputs, and the prediction outputs, to regularly tweak to your hiring decisions and processes.
Step 4: Set up measurement and reporting with analytics recruitment tools
Once you’ve set your predictive analytics machine into motion, the next step is to create a recruitment KPI dashboard to measure the results. This dashboard should be easy-to-use and provide only the most relevant information as it pertains to your core KPIs.
ATS tracking platforms like Recruitee give you customizable options for what information you’d like to see, and how you’d like it to be displayed; be it graphs, pie charts, lists, and so on.
One factor to keep in mind in this stage of predictive analytics in recruitment is the ease by which you can report on your KPIs. Chances are your manager or executives will want to be kept in the loop of what the metrics are saying, and what the results of certain process changes have been. Ensuring your tech stack can provide quick and clear reporting will make your life easier in the long run.
This brings us to the last step (or stage): tracking, measuring, and taking actions.
Step 5: Continuously track and measure success
Seeing and reporting on KPIs means nothing if you’re not making improvements to who and how your recruit. It also means nothing if you set it once and never do any experimentation, or change your recruitment data or reporting metrics.
Predictive analytics in recruitment is inherently a game of change and incremental improvement. It predicts outcomes based on the data it has in front of it, and the outcomes of your actions. To make the most out of these platforms, you should regularly act on your platform’s recommendations, make your own changes and measure the results, and ensure that your data pool is reflective of the outside world.
As mentioned, your predictive analytics model will only be as effective as the data you put into it, and the actions you take with the outputs. Done right, and regularly, predictive analytics in recruitment can yield dramatic results for the quality of your hires and your recruitment processes in general.