Lately, theres a sense I am picking up that the explosion of interest in People Analytics is starting to subside and stabilize. The space is ever so slightly starting to clear.
In the rubble and detritus left behind from this explosion, what can we make out so far?
First, a lot of the fuss that started off the explosion in the first place - predictive analytics - has not had the universal applicability that many claimed it to have. In fact, for most limited-sized organizations and maturing HR functions - and with the onset of GDPR - there is very little scope for successful use of any form of machine learning in an HR context. There are exceptions, but they are few and limited to larger entities.
Equally, the space of HR tech, which saddled itself up alongside people analytics in recent years, is ever so slowly sorting out the wheat from the chaff. Speculative technology cant go on forever and we are finding that much of the technology we have seen in the marketplace in recent years has been speculative. Outside of the really big players who have a highly diversified portfolio stretching outside the HR space, I dont expect many marketplace offerings to survive into the medium term.
The danger of all this, of course, is that this explosion - which has lasted numerous years -has been highly distracting to a function that doesnt need distractions right now. HR has a real job on its hands getting its basics in place and its house in order as we transition into the highly disrupted talent market of the future and plot a path to HR 3.0.
One of those basics is to get the right metrics in place to measure the conditions and dynamics of the workforce. For those who believe that metrics and reporting shouldnt be a part of people analytics - a philosophy I have heard espoused a few times - I would urge you to think more holistically about the purpose and scope of a people analytics function. Would you allow someone to operate on one of your organs if you found out that they didnt understand how that organ worked?
So many HR professionals contact me asking for advice on basic measurement and metrics of their people. Key questions like:
- What are typical metrics used to measure workforce status and dynamics?
- How can I ensure that there is a standard, consistent understanding and use of these metrics across the organization?
- What way should I segment my organization when reporting on these metrics?
I dont claim to know all the answers, and indeed many depend on the specific context of the organization itself, but I want to focus this article on some basic concepts in HR metrics.
The key determinants: purpose, unit, time
Generally, most business metrics depend on three key factors:
- The purpose of the metric - for what reason are we interested in the metric? What problem does it help us solve or what question does it help answer? For example, there is a difference between tracking people for the purpose of ensuring office space and capacity versus for the purpose of understanding the capacity for taking on new work.
- The unit that is to be measured - what metric makes sense for that unit and reflects the role it plays in the organization? For example, tracking days in the office may make sense for field salespeople, but maybe not for software development staff?
- The time period - how do we think about the time period over which the measurement occurs? For example, if you want to know how many people need to have evaluations in the next month, youll probably want to know how many people are employed today, and not the average over the past year.
Narrowing down to HR, the most common purpose differentiation is where the metric is a pure people metric versus a metric that also has a pseudo-financial purpose. The most common time differentiation is the point in time versus reporting period. The unit being measured is usually a population of some form, and different metrics are more or less relevant for different populations
Common HR measurement priorities
Depending on the priorities of the organization, HR metrics usually focus on a number of factors:
- Current state metrics, for example, current headcount or size of the workforce
- Current shape metrics, for example, profile of the workforce by geography, role, seniority, gender, diverse backgrounds, etc.
- Dynamics metrics, for example, arrival, departure, turnover, promotion rates
- Pseudo-financial metrics which are people-related, such as utilization, productivity or compensation
To make the above abstraction more real, here are some common examples of how metrics can be designed around measurement priorities, using the three key determinants mentioned above:
- Headcount is a metric that exists because the business needs to know the number of different human beings currently employed. This commonly means that it is a point in time metric (eg how many employees did we have as at the beginning of the quarter?).
- Size of the workforce can also be a term for headcount if the purpose is to know how many different people are employed, but it can also have a pseudo-financial purpose of understanding the capacity for work in the organization. FTE may be more appropriate than headcount for this purpose, taking into account those on extended leave or part-time schedules. FTE-based capacity measures are more likely to be period-based versus point in time (eg what was our average FTE capacity in the third quarter)?
- Workforce split by group of interest (eg geography/seniority/diversity) is usually simply a breakdown of headcount or FTE capacity into subgroups. Most commonly this is done on a headcount basis, especially if it is for fiduciary reporting requirements (eg EEOC).
- Rates of employee movement (eg arrivals, departures, promotions): Rates exist so that changes and movements can be understood in context. These are usually period-based, but there are subtleties to the specifics of how these can be calculated. Ideally, the numerator reflects the number of people or FTEs that have arrived/departed or been promoted. The denominator can reflect the population at the beginning or end of the period or the average throughout the period. This can cause a great deal of confusion in organizations and, if not well defined and communicated, can result in inconsistent metrics floating around. Small numbers can also result in large, unintuitive rates, particularly if the organizational unit is young/immature.
- Productivity rates can vary massively depending on what the population is. Salespeople may be measured on some sort of revenue metric, while project-based staff may be measured on some sort of timesheet-based metric. Productivity metrics can often need careful discussion and formulation with business leaders. There is also some debate as to whether these sorts of metrics are the purview of HR or should be owned by the finance function.
All metrics need some sort of benchmarking, which allows us to get some sense of how the current metric compares to the broader context that the organization is operating in. The only exception is where the organization or one of its units is very new and where highly volatile metrics are to be expected in the early stages of its growth.
In some cases internal benchmarks are helpful. This is particularly the case if the metrics are best understood in a historical or geographic/business unit context. Looking at organization-wide headcount begs for a comparison with last year or last quarter. Looking at the headcount of a business unit begs for a percentage of the entire organization or some other sense of relative size.
In many cases, external benchmarks are important, but not easy to come by. Understanding how attrition/departure rates compare to your competitors would be a very useful insight but rarely is it possible with any certainty. Diversity representation metrics may be easier to benchmark depending on the geography of interest (eg the US Bureau of Labor Statistics and the Pew Research Center published some good regular statistical reports on workforce representation).
Effective HR metrics and reporting areas are foundational to an HR function as accounting are to a finance function. As you read through this piece, ask yourself these questions:
- How much thought has gone into our current set of HR metrics?
- How well do our metrics serve the entire organization?
- How well understood are these metrics across the organization and how consistent is their use?
If you tick all these boxes, congratulations! If you didnt, maybe its time to spend some time on this and leave the chatbots and algorithms alone for a while!
The post \"The case for good HR metrics\" was first published by Keith McNulty here https://www.linkedin.com/pulse/case-good-hr-metrics-keith-mcnulty/
About Keith McNulty
I lead McKinseys internal People Analytics and Measurement function. Originally I was a Pure Mathematician, then I became a Psychometrician. I am passionate about applying the rigor of both those disciplines to complex people questions. Im also a coding geek and a massive fan of Japanese RPGs.
All opinions expressed are my own and not to be associated with my employer or any other organization I am connected with.