How do you use people analytics effectively?
What does people analytics actually involve?
What are the stages of the people analytics approach?
After last week’s overview, we’re now digging deeper into the people analytics puzzle to answer these questions. We’ll introduce a methodology you can use to develop an actionable people analytics strategy, with examples throughout showing how this can apply in practice.
How to Use People Analytics Effectively: 4 Stages
Question – What Are Your People Problems?
People analytics starts with people. Data can be seductive and noisy, so it’s important to remember that. As we explored last week, there’s a big difference between operational reporting and people analytics. Simply collecting lots of data isn’t enough, if you want to use people analytics effectively.
It’s akin to being told to ‘keep your eyes peeled’ without knowing what for. You’d wind up with an endless amount of information, but filtering through to find something useful would be nearly impossible.
Rather, people analytics becomes powerful when you apply data to your people problems. First, then, you need to know what your people problems are.
This is the observation phrase. You’re not starting with complex algorithms; you’re starting with what’s in front of you.
- What challenges does your organisation face?
- Which issues need addressing?
- What is your organisation trying to achieve?
- How does HR intersect with those objectives?
For example, perhaps your organisation is trying to meet aggressive growth targets and you’re struggling to fulfil hiring quotas. Perhaps merger & acquisition activity is a major strategic priority, and you’re experiencing increased employee attrition. Maybe the Head of Technology is set to retire next year, but the talent pipeline is looking sparse.
According to the re:Work People Analytics Guide, to use people analytics effectively you should ask questions that align with the ‘triple aim framework’:
- 1. Effectiveness
- 2. Efficiency
- 3. Experience
Are your people policies effective? Do they do what you want them to do?
Are they cost- and time-efficient? How could you improve efficiency? How could the qualitative experience of your people policies be improved?
Apply this framework to the people and business observations you’ve made. In the example above, you’re struggling to fulfil hiring quotas to support business growth. To use people analytics effectively, you’ll ask these sorts of questions:
- Are our current recruitment practices effective?
- What are the gaps, and how can we improve them?
- Are we hiring the right people over time?
- How can we make our budget go further?
- How are we perceived as an employer?
- How can we improve the candidate experience?
The problems you face will likely be more multi-faceted than the above, but this gives you an idea. Your people analytics success depends, first and foremost, on asking the right questions. Narrow down exactly what it is you need to know, in order to provide the answers your business needs.
Then it’s time to develop a hypothesis.
Hypothesis – What Do You Need to Test?
A hypothesis is an educated guess, which you can then test. If you want to use people analytics effectively, you really can’t skip this stage.
For instance, you might hypothesise that you’re struggling to meet hiring quotas because your recruitment methods aren’t comprehensive. Perhaps the competitive landscape is fierce.
Maybe your organisation is suffering from culture erosion because of merger & acquisition activity. Your talent pipeline might be sparse because you can’t recruit the right technical talent, because you don’t offer attractive technical training.
Your hypotheses are the bedrock of your story. ‘We believe top technical performers leave our organisation because we’re not offering competitive training and development’.
Then you need to translate your observations into the language of metrics and data. It’s a pairing exercise. ‘If we want to know X, we need to measure Y’. You’re pinpointing which data is relevant to the specific hypothesis you want to test.
In the example above, you’re hypothesising a relationship between attrition and training, so exit data and training data will be most relevant to you.
Exit data/metrics might include:
- Number of voluntary exists
- Exits by department
- Average tenure by department
- Reason for exit by department
Training data/metrics might include:
- Training hours for technical roles
- Training hours for non-technical roles
- Training frequency by role
- Number of training courses available by role
- Training feedback scores by role
The idea is to get as detailed a picture as possible about the training landscape, so you can compare that to your detailed picture of the attrition landscape.
Once you’ve decided which data will allow you to test your hypothesis, it’s time to collect that information.
Experiment – What Does The Data Say?
You might already collect some of this information as default, or you might need to set it up. Some of the examples above are quantitative, while others are qualitative and will demand your active involvement.
For instance, you might choose to set up exit interviews to find out why employees are leaving. You could also decide to issue a questionnaire to all employees during training, for example. This will give a greater depth of insight than relying purely on quantitative data.
This stage takes some time. To use people analytics effectively, the final stage is to draw inferences from your data. This means you need enough data so you can draw meaningful conclusions. A sample size of one won’t get you anywhere!
Exactly how long this takes will depend on your organisation. Hopefully you don’t lose 10 people a month, but if you did you’ll obviously get a much faster set of data than if you lose 1 person a month.
To speed things up, you can break down your hypotheses into smaller, indicative hypotheses. Using the same example, exit data will take some time to collect. However, consistent dissatisfaction ratings from technical performers after training sessions could be indicative that this contributes to attrition.
Conclusion – What Does The Data Mean?
This is the stage where you complete the story, linking your data back to your people problems. This is a vital step, and without it you’ll struggle to use people analytics effectively.
By itself, data is just data. The problem isn’t that we don’t have enough of it – the problem is that HR doesn’t know what to do with it.
If you can turn that data into a story, it becomes meaningful. It slots nicely into a narrative, so we all understand the implications of doing a certain thing a certain way.
Firstly, you’ll use methods of analysis to draw conclusions based on your data. Some of the methods you might use include:
The method you use will depend on what you’re trying to find out. It sounds complicated, but it can be as simple as calculating averages. You don’t need to be a data wizard to use people analytics effectively. Working out cost-per-hire is a form of data analysis.
Your work doesn’t stop here though. As we said above, data is data. If you don’t weave that data into a compelling narrative, your insights will miss the mark. Part of the challenge HR leaders today face is to secure buy-in into these new ideas, in order to impact business decision-making.
There’s a leap here – and HR professionals are responsible for making it happen. Ultimately, you’re using people analytics because you want to achieve something. You’re not collecting data for data’s sake.
Your insight needs to be actionable, and to be actionable it has to be clear, concise and compelling.
Think about the ways you’ll present your findings. This could mean implementing customisable data visualisation dashboards across the business. It could mean a monthly email digest. It could mean a presentation. It could simply mean the way you tell people what’s happening.
Your focus should be on your conclusions, not your data. You’ve aiming to weave together a story based on all the information you’ve collected, painting a picture of the current HR landscape for your business. You’re aiming to deliver short, sharp, impactful messages:
“We’re struggling to build a pipeline of solid technical talent because our attrition is high. This is largely because we’re not offering the high quality of technical training our competitors do, even though our salary bandings are higher. We should reallocate wage funds to develop our training capacity”
The people analytics process starts and ends with people. Phrases like Big Data are thrown around so often that it’s easy to lose sight of that. HR is, and always will be, a people-orientated discipline. We’re simply leveraging data to help us make better, fairer, more accurate, more incisive people decisions. Understanding that is key if you want to use people analytics effectively.