Five ways to get the most out of your HR data

1. Check that you have an adequate sample size

The first thing to do before doing anything with the data, is to make sure that you have an adequate sample size.  The larger the data set the more valuable the information, AND the more you can do with it statistically.  

For example, if you’ve done an employee survey to ask employees if they want communication material sent to their home or to the office, and only hear back from 5%, will that really give you a true answer?  Remember, you want the ‘right’ answer, not just any answer.

2. Sort out good and bad data

As Mark Twain said ‘“If the metrics you are looking at aren’t useful in optimising your strategy - stop looking at them.”

This points out a common mistake, which is to be so busy looking at the data, that we don’t first sort out what is good and what is bad data. Find what data is ‘right’ for you, and discard any data which serves no purpose. For example, let’s say that during your focus group sessions you find that employees prefer hard copies of communication material, but your company is "green" and thus does not allow hard copies. That data would be irrelevant, as implementing more potential waste would conflict with your company’s "green’" strategy.

3. Organise your data

Now that you have adequate data, the next thing to do is to find a way to organise it. Think of it as you would when you organise your closet - the better you organise, the better chance you have finding things in the future. From a data perspective this means thinking of ways to group or section data so that it can be sorted, analysed, etc. This will not only make you effective but save you time in the long run.

4. Look for common themes, patterns and/or relationships

Once you’re comfortable with the data and have done your organisational magic, it’s time to roll up your sleeves and start making some sense of it.  Look for common themes, patterns and/or relationships that emerge from your data. Statisticians call this ‘thematic analysis’, and an example is a statistical tool called regression analysis.  

For example, let’s say you’re asking your employees how much more information they want on the company’s pension plan. You could plot this against their age. You may find patterns that younger employees want to know less than older employees, which would be helpful as you develop your communications plan.

One word of caution: Look out for something called ‘confounding factors’, which is when it looks like there is a correlation but there actually isn’t. Using the pension example, maybe the younger employees want to know less because they recently attended a seminar on this topic, and not because they’re not interested.

5. Present your data in a meaningful way

Last but not least, it’s absolutely critical to present your data and your findings in a meaningful way. This is often overlooked, which means all your hard worked is wasted as yourself and those reviewing the data can make no sense of it. Take a step back from the data and think of what you’re trying to show with the data, and make sure that you present it this way in order to achieve this objective.  

For example, let’s say you’re trying to use your data to get approval to put in place an communications platform. Present the data in a way that points to/validates this solution (e.g. high percentage of employees don’t understand the HR programmes, don’t know where to go to find information, and ultimately don’t value the HR programmes). Data should be able to help you paint a picture in a compelling way, that will guarantee a “yes” from anyone you need approval from.