We should start out by saying that recruiting metrics are important. The quantity of recruiter activity and quality of hires produced by a recruiting team are important figures to understand. Metrics can be used to judge recruiter performance, gauge the quality of employment brand, and develop effective and cost-efficient recruitment marketing.
However, when is all this data too much? When do you move from a smart strategy of examining the recruitment data generated by your team to an the innovation-sucking paralysis of over-examination? When does the cost of examining your recruiting efforts overtake its ultimate utility?
3 Problems with Recruiting Metrics
To understand how metrics can often get in the way of good recruiting efforts, we first have to understand some of the problems with the obtaining and using the data.
- Costly: As recruitment marketing channels became more numerous, gathering intelligence across those channels became very difficult. Organizations now must consider sources of hire from social media, job boards, traditional media, television, search engines, referrals, and their own employee referrals. Gathering the data from these disparate sources into usable and comparable sets is often very difficult. In order to compare these different data sets, it has become necessary to use specialized software for applicant tracking and/or specialized, highly-trained employees. Read: expensive.
- Inaccurate: For most recruiting data sets, there is some human point of input. This could be a recruiter logging an activity call or interview. It could be a hiring manager logging an interview into a VMS or an applicant self-selecting where they found a job online. In each one of these circumstances, there is a very high degree of human error. Very little motivates a candidate to select the proper source of their knowledge and hiring managers aren’t often held responsible for quality recruiting metrics. Finally, individual recruiters are often told to record everything with very little incentive behind this performance – logging activities does not usually prevent a recruiter from being fired. The end result of all of these faulty inputs is terrible, useless data.
- Not Actionable: Some of the recruiting metrics commonly examined have very little follow up and/or consequence. If source of hire was (theoretically) perfectly accurate, a recruiting department can make smart decisions as to which job board to allocate more resources toward. However, consider metrics such as the number of applications per view of a job post. A very low apply-to-view number may indicate a low quality job post, an unappealing employment brand, or unrealistic job requirements. On a per job basis, this metric is quite easy to examine. However, on an aggregate basis, the metric becomes almost worthless. If no one applies to a particular “Scientist IV” position, it may mean it’s because it required a particular niche experience; whereas, in the case of a sales role, perhaps the benefits or compensation model were not detailed or appealing. This is just an example, but you can see that applying universal action or coming up with a plan based on that data is highly problematic.
Consider even some of the most highly regarded, important metrics, such as candidate hire per interview. This data can be very interesting – individual recruiters can be judged on the quality of their candidates and the selection criteria for particular departments can be analyzed. However, it is again a ratio usable mostly on an individual case, because of its wild swings. If you are filling an impossibly hard (or intangible) position, you can and should expect a very low ratio. A “good” hire-to-interview ratio could mean easier positions or indicate higher turnover in the future. It’s difficult to draw any conclusions from this metric that you can use to judge a department as a whole or even the performance of an individual recruiter as compared to another.
A Cost Efficiency Model
Considering that recruitment data is often problematic, hard to come by, un-actionable, and/or expensive, when should you implement these programs? How do you know you are placing too much or too little emphasis on recruiting metrics?
It’s easy to justify costs, either in recruitment software or in specialized human labor, through lofty goals such as improved talent pipelines, topgrading, or employee retention or morale. However, it’s important to understand both the real utility of gathering the data, the validity of that data, and the cost in doing so.
One of the best ways to determine which metrics to follow and/or develop programs and methods to obtain is through a simple cost efficiency model. Recruiting department often quantify only one side of the equation: the cost of a recruiting software purchase or perhaps the cost of a specialized recruiting consultant or training for their staff. It is more rare that department heads will quantify the benefit of the action gleaned from that data.
To understand the cost savings or increased revenue from any program, you must first understand what action you can take from the data. If you can determine that specific action can be taken (such as hiring another recruiter, firing a recruiter, moving to another job board, developing a better employment website), you have to also consider the financial impact of this action.
It’s somewhat easy to say, “If we had this data, we would understand if we should [buy X, reduce X, etc].” It is harder to say “And because we took that action, we will [achieve X revenue, reduce Y cost].”
This is to say, “improving” a commonly watched recruiting metric is not a benefit in and of itself. It is rather the financial consequences of the action taken because of that data that matter.
Recruiting metrics often mean a lot to recruiters and very little from the standpoint of business impact. To justify expense and recruiting program initiatives, it’s important to study which metrics matter most to your organization in terms of financial consequence. You don’t have to feel bad about ignoring the data that is too costly, inaccurate, or impossible to act on. Additionally, if the financial impact and yield from the resultant action will be less than the cost of obtaining and analyzing that recruitment data, you can consider that metric as an expense, not a utility to your organization.