This weekend is labor day weekend in the US, and hence I am bringing some deeper perspectives. An extra day always gives a bit of spare time to refer to more papers. The week was short, as many leaders took vacations both in the US and in Europe. But one of the recruitment leaders asked me a very interesting question.
This question stumped me. I have always believed that I have a decent handling of statistics. But I had not thought about this at all before. The challenge is particularly complicated when the jobs we are hiring do not have a defined metric to evaluate (like a very tactical metric). The roles that we hire have gotten a bit more complicated. In other words, you have access only to the rank order based on criteria you define (ordinal scale) and not a generalized rank order (cardinal scale). This complicates the job of a recruiter. Many recruiters whom I spoke to said, “We just know it” when to push and make an offer. Of course, I was not satisfied with this response. Recruiters have some algorithm they intuitively follow. The quest for me was to define this and document it as much as possible. The issue is, you cannot sample very little, but at the same time, you do not want to miss a great applicant in the name of sampling. With several changes in the labor force, there is an uptick in incoming resumes. So getting this to an “algorithm” will be useful for you. (even if you have your own rule of thumb).
Our research on this took us to several books and papers. A 1960 report on the Secretary problem, published in Scientific American written by Martin Gardner, throws some light into the solution. No one knows the origin of this problem, but the question matches our quest. (When to hire the secretary). In your search, you can fail two ways. Stopping Early or Stopping late (Brian Christian, in his book Algorithms to live by). When you stop too early, you leave the best applicant, undiscovered. When you stop too late, you hold out for a better applicant who many not exists. The mathematical models suggest a simple rule. As the applicant pool grows, the exact place to draw the line between looking and leaping settles around 37% of the pool. Look at the first 37% of the applicants but do not make a hiring decision but post that be ready to leap for anyone better than all those you have seen. Now, this algorithm need not be practiced very precisely. This states that in a hiring process, there is a “look’ and there is a “leap” process. You have to budget a “time based” model to look, and that is often more than just a very few candidates (it can also be tied to the number of positions, talent size in the market, and the number of applicants in your queue). Of course, this can be randomized, and a formula can be arrived to make this entirely bias-free. You can always write to us if we need to design a simple spreadsheet to calculate this. This is what algorithms do. They give a directional sense of solution when solutions appear not to exist. This is both a sampling and queuing theory problem, and many times intuition trumps the solution. We may have stumbled across a fascinating piece of research through this question.
As it is Labor day in the US, I was reflecting on the structural issues facing the Global Labor Market
Here are the five issues that came up at the top
Over the last 60+ years, there has been a dramatic increase in the Women’s hours of work in the developed countries. The ratio of female to male hours increased from below 40% in 1968 to 75% in 2019. But the wage gap seems to have stagnated at around 75% (female to male wage ratio). Even with the growth in the hours, the progress has been slow. This is taken from the Dissertation work done by Kangchul Jo from the London School of Economics. We have to take this seriously and keep pushing the boundaries
Every company is evaluating working from home as a permanent strategy. While I support work from home, the data sets suggested for its favorability are often employee surveys. You can check out a LinkedIn post from an HR leader saying that their employees believe work from home is super productive and other similar sentiments. While this is certainly good news, you do not want long term employment location decisions based on employee surveys at one point in time. The concept of working from an office has evolved over the centuries. (Kepler, who created laws of planetary motion, worked in an office like setup- of course, I cannot validate it). The science of innovation and co-creation is deep and detailed, so we have to take multiple factors into account. Many urban economists have studied this effect in detail. This is a very big issue facing us
An algorithm is a set of steps or a sequence of steps influencing the outcome. Even if your organization has not implemented Robots that you see running around on the campus, they are slowly taking over in hidden formats. From process automation to goal setting to performance standards, we are witnessing a prolific increase in the number of algorithms. Validity, testing, and committees to question the logic should become a critical HR task and HR leadership responsibility. Not focusing on this will build an inherently biased organization in the next decade.
In many ways, every company is becoming a software company. This is a very interesting and weird scenario. The themes of digitization followed by every enterprise demands technical resources. But what enterprises fail to understand is “technical’ resources need not come only from the “Software” world. Throughout these series of emails, I have written about the perils of not reskilling and upskilling. Certain technologies like RPA, Tableau, Python, and other technologies should be learned by everyone. When McDonald’s previous Global CMO Silvia Lagnado took over, she spent three weeks learning the basics of Python. Even though she was not programming it, understanding the basics is essential to ask the right questions
Global unrest across key talent locations is a matter of labor concern. We are tracking all these issues very closely. Very recently, you may have read about the unrest in Belarus. Such disruptions always create talent issues. Hopefully, these issues will be resolved soon