Where people get stuck in analytics

There's no escaping the fact HR analytics has been disappointing. David Creelman identifies the most common reasons why, and how analytics can get better.

Article main image
Apr 3, 2023

For all the success people analytics has had – the reality is that this technology is still…well, rather rather disappointing.

HR professionals were always hoping for dramatic insights. Instead, they seem to have ended up with a bunch more dashboards and reports.

Let’s look at three common problems and my proposed solutions.

Three common problems:

Here are the three most common problems that have bogged down people analytics:

1) The analytics team is too busy to do important work

Analytics has the greatest impact when leaders have an important decision to make and ask the people analytics team for help.

However, this important work isn’t what consumes most of the time of people analytics departments. Instead, they get stuck on producing routine reports and handling complex data management problems so that they have no time to pursue more impactful projects.

2) Leaders do not want to act

Sometimes the analytics team provides some useful evidence to help leaders make a decision but then the leaders simply say, “Can you do more analysis?” This is the famous problem of analysis paralysis.

Another kind of situation is when the analytics team finds an important insight on their own – for example, some insight on remote work – but the leaders simply don’t pay any attention.

In both these cases the issue is not the analysis, it’s leaders unwilling to act on the analysis.

3) The analysis has become messy and inconclusive, so people just give up on it

We often start with a simple question such as “What are the characteristics of employees with high potential?”

When we dig into that question we run into all sorts of difficulties. We find that there is no agreement on who is high potential. We find the data is spotty. We find that the answer varies a lot from role to role. In the end, it all seems too messy and inconclusive to act on.

Three possible solutions:

Here’s how I think we can solve these problems:

1) Separate the analytics consulting group from the analytics operations group

The endless work of just managing reports and dashboards; as well as the difficult work of data management, can easily eat up all the time of your analytics team.

 You need to protect one part of the team, what I call the analytics consulting group, from this endless stream of work.

Have one part of the analytics group handle operations; then have another part act as consultants working with leaders to identify important questions. This is what and brings the power of analytics to bear.

2) Focus on leadership psychology, not data analytics

No matter how good you are at analysis, you won’t have much impact if you don’t understand the psychology that drives leadership behaviors.

You need to develop skills in influence and persuasion. If you thought analytics was a way to get away from the challenges of personality and politics, then I’m sorry to tell you that you were wrongManaging your leaders is a big part of being a successful analytic pro.

3) Re-frame your task by focusing on the key decisions to be made

Problems are almost always more complicated than they first appear.

When the complication risks derailing our analytics project it’s time to narrow the focus.

For instance, if we are interested in “potential” we might ask “Of all the different jobs, which are the one or two where we are most concerned about identifying high potentials?”

But you can also ask: “What decision are we trying to make?” If it’s about making the best promotion decision then ask whether you really need to be measuring potential or if there are other things you can do to make better promotion decisions.

Instead, you might ask “Which two or three promotion decisions are the most critical ones?” Once you narrow down the problem, chances are you will be able to move the analysis forward.


People analytics is maturing and we can learn from the problems of the past.

It’s true many analytics projects get stuck but that’s not inevitable.

We just need to understand why projects are likely to get stuck and deploy tactics to overcome the barriers.