Skip to content

Mature quickly in the HR Analytics journey

Published by orgvue

Analytics is a hot topic and HR analytics is no different. New books are constantly being published. Being “insight driven” is at the core of the CIPD HR practices. New software products and visualisations tools are being launched. The term big data is everywhere (I don’t think HR data is big data, it is human data but that is a topic for another blog). It is all quite a buzz.  But what does it mean? Why is it really so valuable? Where should you start and finish? How far along the journey are you and what does that journey look like?

Lots of questions, perhaps too many for a blog post. But bear with it. The answers are not too onerous, but equally not cut and dried either. The main thrust of the answers is to lay out some of the basics of the HR analytics landscape. What are some of the terms? What are the metrics and pieces of analysis? Why are some of the standard ones flawed, why are they so interesting and where will you end up going with it all?

1 – What does Analytics Mean:

Why not turn to Wikipedia, which defines analytics (and I’m paraphrasing):

“… as the discovery and communication of meaningful patterns in data

… as a quantification of performance

… by relying on the simultaneous application of:

  1. Statistics
  2. Computer programming
  3. Operations research

… with a favour of data visualisation to communicate insight

… and the creation of insight from the past, recommendations that drive action and guidance for decision making.”

In other words, it more far more than just having data or even just insight. It is also about driving action and improving decision making. It is about connections between pieces of data, “meaningful patterns”. It relies not just on statistics and computer programming… but “operations research” (Operations research arrives at optimal or near-optimal solutions to complex decision-making problems).

2- Why is it valuable?

It helps to drive performance and better decision making. Uncover issues, solve them and see the results. The more you can get to seemingly unconnected links the better.

3 – Some basic terminology

Definition of a measure: Measures typically refer to quantitative data, such as number of people employed, number of leavers or % attrition (% leavers), level of absenteeism and so on. In the context of data visualization, measures typically map to the Y axis of a chart.

Definition of a dimension: Dimensions provide structured labelling information to otherwise unordered numeric measures. Dimensions refer to categorical data, such as geography, gender, function or units of time (e.g., day, week, month). Generally, dimensions are used to group quantitative data into useful categories (e.g., number of people employed by function) and typically map to the X axis.

A measure can be turned into a dimension by putting the measure into bins. For instance salary bands or tenure levels.

The human mind likes structure and categories. We like stages and having 7 or fewer things in a list. We also like to know we are making progress and the feeling of success at being able to quantify that progress and tick things off a list.

4 – Where should you start and finish?

As is always said, start simple and build from there. Get the basics right. In other words, “Lay the Foundations”. How many people do you employ? What is the people cost by function (function being a dimension)?

It is useful to create categories along business questions. For instance creating business cases or resourcing as shown in the below table. The last row in the above, the “Dimensions of Analysis,” is a list of ways to slice and dice the metrics.For instance, number of employees by gender or age.

This foundational layer then gives momentum and confidence to push on. Although presented as a set of clearly defined steps in 5 nice chevrons, a hangover from my consulting days, it strictly doesn’t have to be like that. It is clearly possible to take each area (e.g. Resourcing) and drive it through to the end. Equally, you could jump beyond “Monitor the Business” to “Manage Outcomes”.

OrgVue - HR Analytics - Lay the foundationsThe above is not an exhaustive list and not an optimum list by any stretch. There are literally hundreds of “Manage Outcome” metrics. A key set is ratios. Think incidents per employee… or sales growth per sales FTE… or number of patients per nurse… it is endless.

The other issue is 15-20 characters can totally under represent the depth of meaning and not all points on the list are even close  to equal. For instance within “Make Data-Driven Decisions”, “Process cost vs. revenue” can be incredibly profound. What is the “sales acquisition cost vs. average revenue per client?” Entire consulting practices are based on optimising the spans of control. Entire teams focus on absenteeism.

Other metrics are “kind of” flawed. As I pointed out in my previous blog, “Recruitment cost as percent of year 1 salary” can dramatically miss the point. What is the average tenure by recruitment channel or the likelihood of becoming a high performer?

Even if I know the answer, so what? What causes it and can I prove any correlation (let alone causation)? Within HR Analytics, there is so much scope for interesting relationships it just isn’t funny. Is there a link between salary increases or performance or the tenure of managers… versus employee attrition?

What about time? How do any of these metrics change with time? Is there a trend? Are things getting better or worse?

How do I present the information? What pieces of information should be combined on the same dashboard and can people really get their heads around all of this at the same time in any event? Everyone asks for dashboards but do they really just need bog standard simple charts (most of the time they probably do)?

Lots of big points and questions. We are working on making it all simple and achievable. In the coming months the OrgVue team will be writing about different views and examples of HR metrics, dashboards and insights. We will explore a couple of points at a time. It is a debate that won’t and shouldn’t ever end. It is simply fascinating and a great journey to be on.