The use of Data in recruitment is a hot topic right now and I will look to expand on that in a series of posts, but you have to start somewhere, so having taken a new role last summer I’ll start things off by explaining the basic metrics we put in place, why we use them and what we have learnt so far.
Swiftly after I joined, our team took a long look at the processes and reporting before working out what was going to be useful for us. We knew we wanted to get an understanding of the process in a quantifiable manner in order to uncover small problems and get a broad view of trends. In order to find out we needed to be asking the right questions.
What we wanted to know
Data is necessary in order to understand the performance of hiring processes and to be able to improve them. From understanding the dropout ratio by stage, to the average scores by each interviewer to feedback on the process from candidates; if all of this available information is collected correctly it can be used to improve candidate experience, speed up processes and increase the quality of interview delivery. The end goal being the best hire possible for any role in your company, at any time in an efficient manner.
Each company is different, but it appeared our major pain points were the process, the candidate experience, the costs and the lack of management visibility. Consequently, we decided we wanted to know answers to the following questions:
- Were our interviewers consistent in their scoring and questioning?
- Were candidates having a great experience?
- How much was each hire costing the business?
- How quickly was the business growing by department?
- What were the key sources for us at both Primary and Secondary level?
- Were the internal recruiters assessing candidates rigorously enough?
- What were the progression ratios at each stage?
- What reasons did candidates give for withdrawing from the process?
We then started to build out a selection of metrics that allow us to not only improve the internal team performance, but also to provide wider BI to the business. Below, I have set out the metrics we chose and why we chose them:
Total business size: Not strictly a recruiting metric, but we chose it as it gave the management team a nice overview of the overall growth patterns and allowed them to see the scale of hiring as a snapshot. Most importantly it allowed for a quick view of departmental growth within the business as a whole.
Cost per hire by month: We chose this as it’s always a big one in venture backed businesses, as you need to prove team value to the business and your ability to keep down costs to the board. We were able to dramatically decrease the cost of hire from a peak in February, whilst also decreasing overall spend. It tells little in terms of future prediction, but we can now benchmark against the trends in H2.
Hires by source: We chose this to understand what had been the most useful sources in the past, then ideally to inform our choices for the future. We learnt quickly that a combination of advertising, referrals and headhunting was the most effective route to hire, with the simplest route being advertising. It turns out, people wanted to work here and applied!
Candidate Experience: This was an easy one. You always need to track this. If you don’t, you have no idea what the process looks like to your customers. We set this up through a simple Google form passed to every candidate that we engaged with, successful or unsuccessful.
There were 4 criteria:
- Would you recommend Lyst to a friend?
- Were the interviewers communicative and well informed?
- How was the speed of the process?
- How was the communication from the Talent team?
The scoring was a 1-4 scale and there is a noted improvement as both the quantitative and qualitative feedback was taken onboard and put into practice. It is also worth noting that 16 is the maximum score on the following graph.
Aggregate Interview Scores: The aim here was to track our scores in a fairly simplistic manner. We set the score range as 1-4 and defined them as:
- No Hire
- Something Missing
This allowed us to understand at the highest level the decisions being made by our hiring teams. We aimed to hit in the range of 2.5-3.0 as this included hiring manager phone screens (we scored 2.8 for the year). This means that for all the scores given, 53.5% were an immediate hire and 13.2% were listed as amazing. Consequently, the candidate quality throughput is high and again it sets our benchmark for 2016.
Scores by Recruiter / month: The aim was to understand the performance of the internal recruitment team. We did this by tracking the average scores hiring managers were giving to candidates in the process, by recruiter and by month. We could then track alongside the average scores by hiring manager to see if the recruiters needed to be better at screening or the managers needed to be better at interviewing. We tracked this monthly alongside the total number of recorded feedback points to give a broad view of throughput volume.
It’s a little small, but for the purposes of this, green is good, red is bad, basically we hit our proposed target (2.8 average).
Two that got away: We weren’t able to track hiring stage ratios or withdrawal / rejection reasons at the time. The ATS we were using at the time made it very tricky to track these, but we’ll come back to that one as we recently moved to a new ATS.
What we learnt / What we’ll do next
If you take anything away from this, it should be to think about the metrics that work for you. Generic lists of “Top 10 recruitment metrics to track” mean nothing.
I have explained what worked for us and why, but every company will need to find their own (aside from candidate experience, definitely use that one). There are plenty of other metrics we started tracking that we soon stopped and plenty we didn’t we wish that we had started tracking sooner.
It is, however, important to think not only of the recruitment metrics, but also the impact on the wider business as whole. Does your CEO know that you give a great experience to your candidates and can measure it? I promise you, they’ll be happy when they find out. They’ll be even happier when you can increase the number of hires whilst reducing costs…
In addition to this, a lot of the metrics we started out with are historically focused, however we now want to move towards more predictive analytics that will allow us to track hiring performance against peoples performance when they are in a role.