Drowning in Data: 4 Tips for Managing Metrics without Going Under

The problem with too much information is that you think you’re swimming when you’re drowning.

A decade or two ago, one of the biggest challenges organizations faced was getting their hands on useful data in a timely fashion. Then, accessing business data on Monday morning that was current to Friday afternoon presented quite a predicament. But today, even the least tech-savvy company has immediate access to a multitude of graphs and information, available all the time, up to the minute. But while you might think this would create informed management teams and data-based decisions, it hasn’t exactly turned out that way.

On the contrary, the availability of such a variety of data sets has done more to confuse management than to inform it. Now that nearly anyone can view nearly any data set, much of the time that management teams once spent making decisions is now applied to reading and understanding graph after graph after graph. If you have 30 minutes to hear a proposal, and it takes 25 minutes to understand the particular data the presenter is using, that doesn’t leave much time for discussion about how to intelligently act on the proposal.

Your management team’s only option is to pick the right data to review in the first place and present it in a way that informs decisions. These four tips will help.

1. Standardize your views.

The first thing every management team must do is to agree on some standardized data displays. Ideally, each manager should have between three and seven graphs that, together, represent the vast majority of his or her output commitment to the team and the organization. Fewer than three and the graphs are probably too abstract to be useful; more than seven and the team becomes overwhelmed.

True, this requires the team to develop a high level of clarity regarding the output commitments they’re making to each other, but however difficult that may be, it’s a benefit, not a burden. All the data in the world can’t fix a lack of clear thinking regarding one’s intended results. Masking fuzzy goals with fancy plots makes good PowerPoints but bad business.

2. Display forward as well as backward.

Data display is always a retroactive sport. Whether you’re counting units produced, sales made or services rendered, the easiest approach is to create a graph or visual that looks backward at what’s been done. While there’s nothing wrong with looking back, rare is the manager who’s tasked today with producing a result yesterday. All of the work of your organization is in the future, so your data must look forward, too.

Alongside what’s already done, make sure every graph also indicates what’s expected to happen next. That way it addresses what the management team should really be talking about—what they’re going to do now to produce results tomorrow.

3. Then, display forward again.

The thing about management is that it was already pursuing an objective yesterday. When the team gathers to review data and make decisions today, that data shouldn’t only indicate what the team already thought was going to happen—it should also show how its view of the future has changed. Some call that second view of the future a forecast, while labeling the original as the plan. Others refer to them as new forecast versus old forecast, or look-ahead versus target. Whatever your terminology, the key is to design a graph that visually displays your future variance—the extent to which your expectations of the future have changed since you last checked. That way, your graph shows exactly what the management team is supposed to manage.

4. Stick to your views.

You’ve probably already realized it will take some work to arrive at these types of graphs. Not only must you define what’s important to measure, but also answer some difficult questions about how to forecast it, how often to review your forecasts and what constitutes a significant enough future variance to warrant attention. You don’t want to do all this work for a graph in a one-off presentation. Once your management team has arrived at the key graphs—the ones that cover the majority of the output the team members owe the organization—stick with them as a basis for updates, discussions and decision-making.

Run the team to produce that output, and direct team conversations back to the forecasts that predict it. The more the management team asks itself what to do about a clearly defined future variance, the more time it will spend discussing actions it can take today to support or mitigate issues it sees coming tomorrow.

Data informs, yes, but too much data distracts. The conversation in every management team, at every level, should go something like this: “Last time we got together, we thought the following result would happen in the future. Now that we know the future is different than we thought, we have to decide whether to do something about that.”

If your data is helping drive that conversation, it’s helping you and your management team. If it’s not—if it’s part of the vast majority of slick-looking plots and visuals representing interesting information that’s only peripherally related—let it go. In our new data-rich world, that’s the only way to keep from drowning.