Who's afraid of data-driven recruitment?

Recruitment is a number one problem for HR professionals. It's lots of hard work with uncertain outcomes. It takes up all your valuable time. It's too much like a sales job, except that no one ever told you that when entering HR! But imagine if recruitment wasn't so difficult, time-consuming and uncertain. Imagine if you could exert a bit more control over the process. Meet data-driven recruitment - a process designed to make recruitment easier and more comprehensible by using data. What is data-driven recruitment and how does it work?

Data-driven recruitment analytics is generally informed by two separate but connected research tracks. One track relies on recruitments metrics - standardized ways of measuring the health and success of your recruitment processes. For example, the HR team may calculate different metrics from the employee recruitment process data, such as time-to-fill, source-per-hire, offer acceptance rate, hiring velocity and a number of other indicators of efficiency and expertise of your recruitment team.

Some HR analysts don't consider metrics a true people analytics practice - and fair enough, metrics are not insights and don't quite result from weeks of deep analysis. But what then is the point of the lowly metric? Metrics can give us very quick and dirty heads-up that there's a problem somewhere in your recruitment funnel. Metrics won't tell you why there's a problem, but it will serve as a good early problem detection system - of brand strength, of wrong recruitment channels, of bad user experience, of lousy interviewing, etc., depending on where in the recruitment funnel things start to look out of sorts.

To get the best out of your metrics, avoid the 'so what?' trap. This means that you can present your metrics to your team and your leadership, but always make sure that you provide context for understanding what that number really means (You: "Our average time to hire is 34 days". Your CEO: "So what?"). Always benchmark your metrics, to others in the industry, if possible - for example, "Our average time to hire is 34 days, while the industry average in our region is 21. We can improve our time to hire by doing xyz". Also, track metrics over time and across different teams, monthly or at least quarterly - for example, "Our average time to hire is 34 days. This is a 46% increase from the last period. The longest time to hire is in the Product Delivery department, on average 45 days, let's take a look at what's happening."

Some of the more popular recruitment metrics include:

Time-to-fill: time elapsed before the position is filled, starting from the date of requisition to the date the candidate accepted the job offer. SHRM reports a benchmark of 42 days average, but this can change a lot from industry to industry. Time-to-fill tells you a lot about the state of the job market and the overall speed of the hiring process. It has important budget implications - the longer it is, the more it raises cost per hire.

Time-to-hire: time elapsed before the position is filled, starting from when a candidate entered the pipeline (sourcing, application, etc.). Time-to-hire reflects on the effectiveness of the recruitment team - how quickly can decisions be made on a given candidate. Long time-to-hire is one of the biggest threats to good recruitment, because top candidates can be grabbed in as little as 10 days. It is also the number one candidates' complaint in all of our prior research - that the hiring process simply "takes too long".

Offer acceptance/rejection rates: percentage of applicants who officially accepted/rejected the offer after it was made. High acceptance rates indicate competitive offers and generally positive candidate experience. Conversely, consistently high rejection rates point to a problem with the recruitment strategy: non-competitive salaries, insufficiently strong brand, lack of communication and clear expectations and other issues.

New hire retention rate: what proportion of your new hires stayed with the organization during the time period you're interested in (12-24 months usually). To do the calculation, just divide the number of hires that stayed with the total number of new hires in the period and multiply by 100. If at all possible, take a look at the distribution of voluntary departures - for example, if most departures take place in the first few months, there may be something wrong with expectation setting, organizational culture, interpersonal relationships, etc. On the other hand, if most of them leave closer to the 2 year mark, there may be insufficient career opportunities or pay increases.

Application abandonment rate: percentage of applicants who abandon a job application after they started filling it out. You can calculate this rate from your website analytics data quite easily and pinpoint problem spots where you lose most of your candidates. Globally, this rate is staggeringly high, from 60% to 90% - this shows that the online application process is ripe for a redesign.

Applicant conversion rate: the percentage of unique online visitors to job pages (job ads on career site, for instance) who turn into applicants. For this metric, you'll need the data from the job board provider or career site analytics. While the conversion path is not always linear, a benchmark to aim for is around 10%.

Total applicant reach: sum of all potential applicants reached by job openings, across all channels. This means total unique visits to career pages, plus job board visits, plus social media reach (unique people who saw your content on social media). Additionally, you can include in this metric the recruiters' direct outreach, or the number of face-to-face event visitors. Because of potential overlaps, total applicant reach is a rough estimate, not an exercise in precision, but it can be a useful indicator over time of the exposure your job ads are getting.

Applicant quality: rate of the number of applicants who are selected for the interview in relation to the total number of applicants. You would calculate this rate by dividing selected applicants by total number of applicants and multiply by 100. The benchmark to aim for is 75% - if it's lower, that means you are getting lots of unqualified applicants to apply for the job or that something is wrong with your screening process.

Source-per-hire: Our favorite metric here at the People Analytics Hub. This metric shows you where your candidates are coming from - for instance, 60% of your hired candidates came from internal referrals, 20% from external recruiters, 10% from job board ads and 10% from the recruiting event you've organized. You can dig deeper into this metric by looking at where the most quality hires come from (by using performance data, for example) and invest more heavily into that particular channel.

Tracking metrics over time is important. Not only that you'll be able to spot certain trends, but they are also a great baseline tool to judge the success of your recruitment pilot projects or changes. For instance, let's say that job board ads generate on average 1000 visits, but gave you only 5 applications (job board applicant conversion rate of 0.5%), which is pretty low. Then let's say that based on this metric, you decide to modernize your job ads, to write better copy, do better branding, add some photos or videos, shorten the application. By tracking this metric in the next few months, assuming nothing else significantly changed, you can estimate the extent to which your change efforts hit the mark.

Generating insights

The second research track relies on generating insights about your particular target audience. You can do this by using your own employees (internally), but also don't be afraid to reach out to external members of your target group. Why? While your internal employees represent a convenient sample, they are also a biased bunch (presuming they like working at your company). External, or potential candidates can give you a much needed counter-weight in this situation. So, what are we looking to find out?

Candidate experience is the number one area of interest here. Candidate experience refers to everything your candidate think and feels about your company from the very first moment they become aware of the job opening to the moment they get hired (or not) for the position they were applying for. Measuring candidate experience means identifying touchpoints (points at which your candidates interact with your company, either directly with your company's employees or through exposure to your company's employer brand) and figuring out which touchpoints create positive and which negative experiences for the candidate. For example, when checking employer review sites, do your candidates form a positive or negative impression of the company? When at your career site, are they experiencing a user-friendly, intuitive and informative journey, or do they have a confusing, overly complicated or frustrating user experience? How do your candidates feel about the interview process - was it respectful, fair, accomodating or did they go home afterwards and told every one of their friends and colleagues what a lousy experience they had?

Candidate experience is still in its infancy, which provides your team with a great competitive advantage if you take your candidates' experience seriously. A recent survey of the Fortune 500 companies (State of TRM Report) paints a bleak picture: 92% of career sites do not have a way for candidates to apply with social media credentials, 59% do not provide information on why candidates should choose to work in their company, only 18% include videos/photos on their career site, and the bottom 17 have no career sites at all.

Recruitment funnel

One way to easily conceptualize a candidate's experience is to frame it as a 'funnel'. Recruiting funnels, even though quite simplistic in comparison to real life candidates' trajectories, can be a useful representation of the stages your candidates have to go through to get a job at a company.

Figure 1. Recruitment "funnel"

At the top of the funnel is Awarenes - at this stage, applicants become aware of the job opening. Here, it's very useful to know where your applicants first heard about the job or the company and what impressions they've made. In your research, always ask your applicants what kind of questions or dilemmas did they have at each stage and how well were they answered. These answers will guide you to create the right kind of content for each of the stages. In the awareness stage, employer branding is key - your brand messaging should clearly spell out what is it that makes your company more attractive or different from other employers in the field.

Second stage is Consideration. In this stage, candidates do their own research - they visit employer review sites, company pages, and ask their friends, family, colleagues and ex-colleagues for opinions. The main question they have at this stage may be "What is it really like to work at this company? Who will my colleagues be? What's the culture like?". One outcome of your candidate research here is to come up with a list of influencers for each of your target groups - for example, your junior candidates may rely more on the opinions they find online, but senior candidates may find their colleagues' opinions more trustworthy - this data can direct you to the right channels and shape your recruitment messaging for best possible impact. Another very fruitful research avenue is career site user experience, as it will reveal a host of employer brand perception issues.

Third stage is Application. Metrics may tell you that there's a big drop off in this part of the funnel, but as we said, those numbers won't tell you why. You can rely on research with new hires to provide some qualitative data on the application process (as it is still fresh in their minds) and adjust things accordingly. You may hold a 'redesign' workshop where your employees can suggest improvements. Alternatively, you can also sample some potential candidates through a simulation task and ask for quick feedback on the application - is it too long? is it repetitive? is it mobile-friendly? did you get a 'thank you' note? did you need one?, etc.

Forth stage is Selection. This stage is composed of screening, assessment tests and interviewing. Many metrics can be followed at this stage, but the best insights will be gained by talking to or surveying your candidates about their experiences with these selection steps. Providing an opportunity for candidates to give feedback on these processes serves 2 purposes: one, it allows you to collect useful data on how to improve, and two, it allows candidates to be heard and if needed, to 'vent' their frustrations. If given an opportunity to do that directly, they are less likely to leave bad reviews on employer review sites.

Fifth stage is Hiring. This is where recruiters finalize the process by hiring chosen candidates. The most important thing to understand here is offer rejection - why some chosen candidates did not accept the offer. If more than 10% reject offers, you need to dig deeper by conducting short interviews/debriefs with that specific group. Rejected candidates can be a great source of data too, especially because you want to keep them in the future talent pipeline too. Devise a follow-up protocol for all candidates - for example, a brief survey measuring different dimensions of experience, as well as the obligatory Net Promoter Score (Figure 2) for recruitment. General rule of thumb: if you establish trust and transparency throughout the funnel, candidates are more likely to respond to the follow-up questions about how your hiring process might improve.

Figure 2: Net promoter score for recruitment ("Based on your recruitment experience, how likely are you to recommend us as an employer to a friend or colleague?")

Some important pieces of advice we give our clients as they embark on the introduction of data-driven recruitment:

1. Make sure that you have a well-defined system in place where your recruiters will enter the necessary data, in a systematic and timely way according to the specific data plan. Recruitment software is a fantastic investment - aside from keeping everything well organized, most recruitment software currently on the market are equipped with some cool data reporting capabilities.

2. Do not underestimate the power of qualitative data! Many HR analysts that we've worked with are terrified at the mention of analyzing a hundred pages of interview or focus group data. Sure, it is time-consuming, but the payoff is worth it. Deep insight can only come from deep diving into your target groups' perceptions, thoughts and feelings.

3. Take the time to segment your target group. Your workforce is not homogenous - different groups of employees have different needs and expectations of their recruitment experience. You can segment based on experience level, type of job, department, or some other characteristic you find important. Sometimes it's wise to go beyond segmentation into full-on personalization, especially when it comes to candidate outreach messages and feedback.

4. Data-driven recruitment is a group effort. Engage your stakeholders, engage your employees, engage your candidates. Build trust and transparency throughout the process and listen to what people in all of these roles have to say. Create internal events, give internal talks, teach people about the importance of data in designing a competitive recruitment process.