The foundation of recruiting performance has been built historically on three core business metrics:
1) Cost Per Hire = Can you recruit and do it with optimal financial investment?
2) Quality of Hire = Can you recruit an optimal or better performer?
3) Time to Fill = Can you fill the position quickly?
For this discussion I am going to concentrate on the third one, time to fill, which is historically a calculation from the clock starting once the business comes to recruiting with a need, and then stops once the candidate is hired/or onboarded. I want to share with you the journey that the Avanade team and myself have gone on, and how we arrived at the conclusions that it was time to blow up the time-to-fill metric.
If you speak to a business leader they will tell you that one of the key indicators that they need to see from a recruiting function is how quickly we are filling the business demand. Simply, most companies primary existence is to grow, and growth is fueled by people. If you cannot hire the people quickly, then this impacts growth and revenue particularly in a professional services environment. Last time I checked, most companies exist to make money.
I found a few fundamental flaws in the time to fill metric:
It only really shows the beginning and the end date/time stamps as an average of multiple roles, and as we all know in recruiting, there are a lot of things that can go wrong in-between the moment you open a position and when you finally fill it. In short, the metric really does not help identify the “why” vs. “what.”
Is the reason that a role is taking so long to fill because of the identification and attraction of that talent?
Is it because of how long it takes the business to assess, interview, and hire the candidate?
Does the metric account for proactive recruiting approaches where you pipeline talent ahead of the actual demand vs. reacting to just-in-time demand if an employee just resigned?
The real issue is if we want to add business value then we must go beyond capturing backward-looking data and move to analytical reporting where we can identify the outliers of peak over- or under-performance. So here is a visual of what we did.
TTH (Time to Hire) = Meeting the needs of different business realities
As we get more proactive as a recruiting function, we start to get line of sight around business demand that can be a multiple of quarters further out in our fiscal year. If the business tells us of their Q4 demand in Q1 of the fiscal year, we can start proactively identifying, attracting, and pre-screening candidates prior to that demand needing to be filled in Q4. Historically time to fill would have meant that a recruiter opened up a requisition in Q1 and the clock started ticking; when the business could actually hire that person based off business demand in Q4, the clock would stop. This meant that the time to hire would be nearly a year long.
But hold on a second: in this particular situation recruiting and the business worked proactively together, but the time metric would be telling a different story. So by simple changing the approach and creating an Open Demand req where the Req Target Hire date would be in the first week of Q4 (which is when the business says they can hire that person), then if we proactively source, screen, and get the business to proactively interview just prior to the actual hire date, then in fact the new time metric could reflect less than seven days against business demand.
When speaking and validating this with business leaders they agree that this approach is a true reflection of their business reality and recruiting’s role in optimizing the process.
This works just as well when you deal with a just-in-time position that needs to be filled if an employee left the organization. The clock starts when the business says they need the role filled (now) and the clock stops when the candidate is hired. In both approaches we can now analyze and separate by job family, geography, business unit, etc., on outliers for both proactive vs. reactive areas for optimization.
But to get to the real essence of an opportunity of “why” something is not as optimal as it could/should be we need to look at:
STI (Source to Interview) & ITH (Interview to Hire) metrics
As we all know in recruiting sometimes a role is difficult to fill because the type of candidate you are trying to find is not readily available in the market and takes time to identify and attract. Or there are times that the reason the role is taking too long to fill is because the business is taking too long to engage, interview, and make a hiring decision. In the old time-t0-hire approach where one metric and approach is designed to look at the entirety of the process and an average, this can become difficult to compartmentalize where the problems actually are. Is it a sourcing/recruiting and attraction problem, or is it a problem once the business takes over?
In short, by tracking source to interview, we can see if certain roles in the business require longer runways of identification and attraction (the date the candidate record is created in the database). In turn, we can go back to the business and explain that by giving recruiting a greater advanced view on this type of demand is more critical than other roles, we can proactively start building relationships with key talent well in advance of the need.
The other recruiting industry trend that we are all seeing is that the future of talent acquisition will be more focused on the building of longer-term relationships with talent vs. the identification of talent. By tracking the candidate (or lead … name/title/company) date in a database we will find instances (which I have already over the last five years) that some roles end up getting filled by candidates who were first identified and the relationship nurtured over 18 months ago. The more proactive a talent acquisition function becomes with workforce planning and succession planning, the more this scenario comes into play and is critical to the growth of a business.
ITH (Interview to Hire)
By tracking interview to hire, we can quickly identify the “why” outliers for both positive and negative reasons against roles, hiring managers, job families, etc.
Let’s say two hiring managers with the same business unit responsibilities differ on interview to hire, and when you dig into the “why” you find that one effectively hires candidates with three business interviews vs. the other does seven business interviews. This might be OK, but it could also be an opportunity to take the best practices of one hiring manager and apply them to another similar manager.
Another example is where one geographical region takes 3x as long on their interview-to-hire metric than another. When you dig into the “why” you find that the business interviews end up canceling or consistently rescheduling candidate interviews significantly more than the average –and the reason why. Once again, you now have the “why” that allows the conversation to be focused on the opportunity for improvement specifically to the situation vs. historically being challenged to identify the reason because it is buried in a broad average that includes both sourcing, recruiting, and the business.
Important Operational Note: I drew a line in the sand that stopped the calculation at when the candidate accepts the offer (hired). But we still can track when the candidate starts (turns up on the job) at the company, given countries have very different criteria around notice periods, which is not globally universal.
We recently deployed this new framework within the organization and we will be looking very carefully at the data in the ensuing months and years so we can not only start to determine benchmarks on these new metrics by company, country, job families, and hiring managers, but most importantly use the information to help drive greater performance for recruiting and the business.