Apart from croissants, oats, fruits, coffee and juice, here are the takeaways from our breakfast briefing with Virgin Media : Using Predictive Analytics to Understand Employee Behaviour.
Most companies are exploring the use of predictive analytics, with some being further on the journey than others. In a nutshell, predictive analysis can help you identify problems before they happen. You can use historical data to predict behaviours that are likely to happen, and then put measures in place to change them.
The possibilities of using this type of analysis is endless. Use it to help identify the cause of absence, stress and attrition, among other important factors. When embarking on a predictive analytics project, it is important to take the following points into consideration:
Aretruly predictive of employee behaviour?
Engagement surveys are often the starting point for most people analytics projects, but are they always useful? Yes and no. They have their place but are often only a snapshot of what’s happening at the moment of the survey.
Can they be improved? Definitely, our research into the validity of hundreds of different employee engagement survey questions shows that very few are actually predictive of future employee behaviour.
When collecting baseline data – define what you are trying to measure.
It’s better to pick a particular question you want to answer, and solve that, rather than try to solve lots of things at once. Find realistic metrics, use data you’ve got access to, and get buy-in from leaders by giving them specific outcomes.
For example, you can define engagement however suits you best, and ask the questions that will get the most meaningful results for you. Ask questions that are predictive of what you are trying to measure for example – absence, stress, growth or productivity.
People analytics must take into account rich, qualitative analysis
For a people analytics project to be successful, the People Insights team must go beyond the numbers and find out the causes and solutions to real people issues. The only way to do this is by understanding the rich open text feedback in your employee data.
Quantitative statistics give you an understanding of the problem. Qualitative analysis adds your employees’ experience, empathy and common sense into the equation to provide context and help uncover the actions that will have the greatest impact.
So, what do you need to know if you’re preparing to kick off your first predictive analytics project in your business? We thought it would be useful to give you a checklist to get up and running. Here are 5 steps to get started:
1. Define the outcomes that are meaningful for your business.
Set out the broad aims for your project and align it with your business requirements. Our customer, Virgin Media, explained that everything they do in their People Insight Team must be related to their business mission. In their case it is their pursuit of the ‘North Star’ as defined by Sir Richard Branson:
“My philosophy has always been, if you can put staff first, your customer second and shareholders third, effectively, in the end, the shareholders do well, the customers do better, and you are happy.”
Therefore, their main aim is to grow through advocacy and become the most recommended brand by their people and their customers. To make this meaningful for the People Insights Team they measure this goal through attrition for others it could have been recruitment referrals or another metric.
This becomes the starting point for any project.
2. Identify the drivers you can focus on
Next, understand the drivers that might be impacting upon what you are looking to measure. Examples of this are leadership, tools, purpose and values, but all of these need to be measurable. If you haven’t got data on these topics already you can look to survey your employees to understand these additional dimensions. Qlearsite has created our own framework to understand each dimension – Qlearsite Organisational Fitness™, as illustrated below:
3. Understand how these focus areas directly link with outcomes and business metrics
Using predictive analytics you can find correlations between these drivers and the business metrics. By analysing the correlations between these two, you can begin to understand whether or not changing certain parts of your organisation will have a material impact on your overall business metrics.
4. What data sources do you have available
Consider what People data you have available. As well as your typical Employee Register file, review what other data do you have access to. For example, learning data to assess how successful training has been.
5. How will you continue to measure the new metrics?
By building these metrics into your business scorecards, you can keep a regular temperature check on how you are doing. This can be shared amongst the People Leadership Team or beyond to the Executive team, giving transparency as to how you are progressing on the metrics that matter to your business.