Data Sensemaking Requires Time and Attention

Data sensemaking requires skill augmented by good technologies. Even though data sensemaking skills can be developed by almost anyone, they can only be acquired through sustained effort to learn the relevant concepts, principles, and practices, which takes time. This hard fact isn’t as appealing as the fantasy that technologies can turn us into data analysts over night. It is largely because technology vendors have been selling us this myth that we have been trapped in the “data age” but have not yet managed to enter the “information age.” For 35 years I’ve been involved in decision support, data warehousing, business intelligence, analytics, data science—call it what you will—and I find it appalling that, despite all the hype about the information age, we’ve made little progress in deriving value from data. Until we accept the fact that data sensemaking and the decisions that it informs cannot succeed without skill, we’ll remain stuck. Besides skill augmented by good technologies, there are two other prerequisites for successful data sensemaking that are usually ignored but deserve mention: time and attention.

Only those data analysts who are given time to explore and analyze data thoughtfully and thoroughly are consistently successful. Most of the great data sensemaking discoveries that we hear about are the result of work by those rare data analysts whose organizations have given them time to do their jobs well. In contrast, most data analysts are either churning out responses—usually in the form of reports—to a long list of “urgent” requests at a breakneck pace or have other jobs that take most of their time, so they squeeze their data sensemaking activities in here and there whenever they can. This is the norm because few organizations have realized that getting real value from data doesn’t just happen “techno-magically,” as they’ve been led by vendors to expect. It takes time to learn and develop data sensemaking skills and it continues to take time to apply those skills each day. This is because data sensemaking involves analytical thinking and analytical thinking takes time. Technologies can assist by doing fast calculations and other forms of data processing, but the thinking that’s required is slow.

Effective data sensemaking also involves attention. This is one of the requirements for rich thinking that has become more difficult to achieve in recent years. Our lives have become increasingly disrupted by the constant demands of the “persistently connected” technologies that we’ve adopted. For how many minutes can you become lost in concentration without being pulled out of your intellectual reverie by a beep, ding, ringtone, vibration, or the sudden appearance of an alert on your screen?  This is not what’s usually meant by disruptive technologies and it certainly isn’t a desired effect. Whenever our attention is pulled away from a data sensemaking task, it takes time and effort to get it back, and much can be lost in the meantime.

Data analysts need an environment that supports concentration without interruption for extended periods of time. The open floor plans that many organizations have been experimenting with to promote collaboration (and, let’s face it, to also save money) are perhaps appropriate for some jobs, but not for data sensemaking. Most of us need to shut the door, turn off all possible sources of interruption, sit in a comfortable chair, and think attentively for long periods of time. I never had a chance to meet the celebrated Princeton statistician John Tukey before he died, but I’ve heard from friends who did that when he took on a consulting job for a client, he would begin with a meeting to discuss the problem and he would then cloister himself in his hotel room for a couple of days before emerging with a solution. Imagine Tukey trying to navigate the disruptive environment of the modern workplace. If he had, perhaps we would have never heard of him. Perhaps the box plot and his many other analytical inventions would not have graced our world. He knew that data sensemaking required attention and he structured his environment to provide it. We must do the same, which means that our employers must recognize the need and support it.

I recently gave a keynote presentation at the University of Cincinnati’s Analytics Summit, and when I mentioned this rarely addressed need for time and attention, you should have seen the heads nodding appreciatively throughout the ballroom. Their eyes brimmed with gratitude for my recognition that the conditions of their work do not match the professed commitment of their organizations to analytics. Claiming to embrace analytics is a far cry from the investment that must be made to fulfill this claim.

Is your organization analytically savvy? Does it exhibit a culture of analysis? If your answer is “Yes” but your organization doesn’t give you adequate time and a work environment fit for focused attention, you’re setting the bar too low. Imagine how much better a data sensemaker you could be with more time and attention.

Take care,


3 Comments on “Data Sensemaking Requires Time and Attention”

By Marty Gierke. June 10th, 2015 at 4:42 am

Well said Steve. Couple the lack of private space and constant disruptions with ever-tighter budgets and increasing impatience of those looking for magic pill solutions to tough questions and you have a recipe for mediocrity at best. Companies who profess to be “data driven” are misguided if they don’t back up that claim with real evidence that they respect the craft.

By Kris Erickson. June 11th, 2015 at 8:01 am

Completely agree.

I believe there as been a regression (no pun intended) in data sense making due to a convergence of a two factors recently. Number one is that the great recession cut many positions at a time when data was growing tremendously. This resulted in higher demands per person. The removal of an admin who isn’t doing data does have an effect on those who do.

Number two, is that the proliferation of systems generating data keeps increasing. If an IT team has X people and they produce a new system each year, then after 10 years there are 10 systems doing 10 useful things and generating meaningful data. In addition, suppose the business team has the same Y number of heads (or a growth of +~5%/year). There is now a geometric increase in complexity and possible interactions between datasets, but a near-linear increase in people to understand that data.

As a consequence most of the time I’m doing glorified counting (#, ratio, %s) instead of true science.

By derek. June 16th, 2015 at 6:09 am

What can I say but that I agree with you and all the commenters? So many organizations in my experience will do anything to support the ideal of the Intelligent Business, except give the staff on their payroll a minute’s peace. No wonder academia has a reputation for intelligent insight, and a simultaneous reputation for letting their people noodle around seemingly forever without being “productive” (I’m aware the reality is often not like that)

Some days I feel like a printer driver made out of meat: “kindly provide us with a table of these values, no need to think about them, we’ll do that, you just get on complying with the next demand for raw data.”

Leave a Reply