There is an uncontested belief that HR should produce dashboards. Often the ability to produce dashboards is one of the justifications for spending a lot of money on HR software. Given this belief, HR had better produce those dashboards. If you’re not doing it, well someone will be wondering what’s wrong.
The problem is that most HR dashboards are not used. In fact, the most important dashboard metric is: “How often people use the dashboard to make a decision”. That number is often very small. I recently ran several HR analytics workshops in Asia; the participants had great stories of how they’d stealthily discontinued preparing some time-consuming report and that it was years before anyone noticed. One HR leader even recalled a manager who had added a zero to the numbers (e.g. changed “1000 staff” to “10,000 staff”) to see if anyone questioned where the extra 9000 staff had come from—again, no one said a word.
So you may have to produce dashboards because everyone expects you to, but you have a right to fear it will be a waste of time.
How do we fix this?
HR, rightly concerned about the value of dashboards, typically starts with the question “What should we report?” Here is lesson number one: the starting point is to report nothing at all. Don’t include anything until you’ve a clear and important reason for generating a number.
Setting such a high bar for HR metrics is a big change in perspective. “This might be interesting” or “This would be good to know” or even “Managers asked for this” are not good enough reasons to include a metric in a dashboard. They are not good enough reasons because we know from experience that often people don’t use the data and that producing accurate reports is expensive.
A good reason for including a metric in the dashboard would be along the lines of “Store managers need weekly reports of how much has been spent on overtime so that they can adjust the schedule such that they don’t go over budget.” What we have in this example is a clear action driven by data. That’s what we’re looking for, and that is the key to resolving your dashboard headaches.
When it’s clear how a number will drive an action then we can be confident it belongs in the dashboard. Don’t feel obliged to include anything that falls short of that criterion.
Start by focusing on frequency
It will make your life simpler if you start by focusing on how frequently a given metric might drive action.
Let’s imagine you have a monthly HR dashboard; does it make sense to include data from your engagement survey that is run once every two years? Clearly not. If you need to create dashboards, then plan to have different ones for daily, monthly and annual metrics (or whatever time periods suit the cadence of your business.)
On a daily dashboard there should only be numbers that change in a significant way on a daily basis. Very few HR metrics fit that criteria and it immediately leads us to the realization that a daily dashboard may not be needed at all; and a weekly or monthly dashboard may only be needed by a few managers. This is good, it means we won’t do a lot of unnecessary work providing data that’s not having an impact on action.
With annual data we have a bit of a crisis because if we’re thinking in terms of “using data to drive important actions”, but we’re only dealing with things that can wait until year end—well, what is highly important, but non-urgent? The answer tends to be in setting priorities. For example, HR might want to have a look each year into which areas have the biggest problems in quality of hire. A dashboard with quality of hire data would help focus attention where it’s needed.
From standard cyclical to on-demand dashboards
In the last example, you probably were thinking “Shouldn’t we look at quality of hire issues when it feels relevant, rather than wait for a set of standard metrics to show up at year end?” This brings me to one last point: reports and dashboards date from a world where a central team had to push information to users. We are now in a world where users can pull the information they need, when they need it. Increasingly they can do so by themselves thanks to self-service technology. Pulling information as needed, rather than pushing information we suspect someone wants, is the future.
This is yet another hint that we should be ditching the idea that HR should be producing a lot of dashboards. There will be cases when managers do want regular reports; for example, a talent acquisition leader may want a report (or at least a warning light) on the number of open positions so that they can decide if they need to intervene to speed things up. But let’s only produce dashboards where the need is clear and specific.
I’ve had to do a fair bit of self-editing so I don’t sound too cynical. Dashboards can be useful and it’d be foolish to throw them out. However, the situation HR has found itself in is that it’s expected to produce dashboards even when there is no clear value in doing so. When people discover there is no value they don’t say “Oh, never mind, there is only limited need for HR dashboards”. Instead they say, “Well, you’ve done something wrong, come up with a better dashboard”; and thus HR ends up spending lots of time and energy on an impossible quest.
Remember, your starting point is that the dashboard should have no metrics on it. If someone suggests a metric, then make sure they have a clear and impressive business case for how having that metric will drive an important action. If you end up with no dashboards at all, that might not be such a bad thing.
Sidenote: Check out David’s previous post on how to deal with analysis paralysis as well.
David Creelman & Bob Nagar
David Creelman is a thought leader on HR and people analytics. His unconventional insights help organizations get unstuck and deliver more value from analytics at much less cost.
Bob Nagar works with Creelman Research providing an unusual mix of expertise based on his successful career running an IT business as well as working on organizational development and executive coaching.