The Disappointment of Data

Driving the other day, I saw the following electronic highway sign:

Seattle:
Rte. I-90 19 mins
Rte. 520 14 mins

There are only two bridges from Bellevue to Seattle, so I had to choose one of these routes. No-brainer, right? Take Rte. 520 and save five minutes.

Not so fast, so to speak. Seattle’s not a small town. Both these routes lead there, but they leave drivers in very different places. Does “Seattle” refer to where each road enters the city, or to a specific spot? If the latter, where? Depending on where I want to go in Seattle, I-90 may indeed be the faster route. Or maybe not.

In other words, this sign contains far less information than is apparent on the surface. Worse, it’s not at all clear how to act on whatever information I might glean.

It contains data, not information.

The Data Dump

Some managers – project managers, people managers, business managers – are in love with data. “Give me data,” they cry, and their teams scurry around to dredge up numbers from all aspects of their work. They measure the measurable, and develop substitute metrics for the unmeasurable – or make up numbers. They hound others for data. They count what counts, and count what doesn’t count. They toss factoids galore into the manager’s blender; the manager presses “puree,” and out comes… mush. 

But the manager doesn’t know it’s mush, or can’t distinguish mush from a strawberry smoothie. She has data to cover every base, and cover her behind as well.

But data devoid of context is just that, data. It’s not information. It doesn’t drive informed decisions, or foster creative solutions. Neither managers nor team members can act efficiently on data alone, no matter how much of it they pour from that blender. 

When a manager says, “I’m data driven,” or “this is a data-driven organization,” do some driving yourself in the opposite direction. Or get used to disappointment.

Let’s go back to that all-data-no-information highway sign. Given the same three lines of about 18 characters each, the sign could easily have read:

Space Needle Rte 520
Waterfront Rte I-90
Conven Ctr Rte 520

There’s not as much raw data, but there is now information I can act on. If one bridge is blocked by an accident, as too often happens, the sign can offer that information instead; otherwise, both commuters and visitors can benefit when providers transform the disappointment of data into useful information. 

A simple example of this transformation is the unloved status report.

Status Reports

It’s said that people detest status reports because they so clearly demonstrate the lack of progress.

Unfortunately, most status reports demonstrate nothing at all. They take too long to write, they’re often out of date or less than honest, and they offer little to act on. In other words, they contain data but little information.

The 3×3 (“three by three”) status report addresses most of these issues. It can’t cure dishonesty or directly jumpstart progress where there is none, but it does capture and transmit useful information in a limited time. They take little time to write and to grasp.

A 3×3 contains three sets of up to three bullet points each:

• What I/we did since the last 3×3 
• What I/we will do before the next 3×3
• Issues I/we need help with or want to raise an alert on

Each bullet point is limited to a single line, and none of the three categories can have more than three bullet points. Generally I suggest doing 3×3 reports either weekly or fortnightly, though there are teams on fast-moving projects that do them daily or thrice weekly at a morning “war room” meeting.

If it takes time to recall what you’ve done in creating a 3×3, you probably haven’t accomplished much. Likewise, being unsure or unwilling to commit to work in the next few days isn’t a good sign. And don’t be afraid to ask for help or alert the reader to a potential issue.

As a manager, I used 3x3s to keep current with multiple teams and projects, to catch shifting priorities, and of course to spot roadblocks I could help remove. 

I would compare the “three things I did” with the previous iteration of “three things I will do.” Disparities sometimes mean problems have been uncovered or progress has been slower than hoped or promised; however, often they reveal that someone – quite possibly me – is resetting priorities, whether intentionally or accidentally. I remained particularly alert for situations where a team member understood a change in priorities but hadn’t confirmed the reality of the change – e.g., taking an offhand comment as a directive.

Be at least a little suspicious if there is never anything in the third grouping, “issues.” It’s not uncommon for there to be a 3×3 period with no issues. However, if a given person or team almost never reports an issue, that’s a sign that there may be an underlying team-dynamics problem.

I didn’t necessarily ask that a 3×3 be presented as a document or EMail. As a department manager, I’d meet with my direct reports in weekly or fortnightly one-on-one sessions and simply ask, “What’s gone well this week? What’s not going so well? What can I help you with? What’s on your plate for next week?” (Together, the second and third questions represent the “issues” part of the 3×3 in conversational form.) 

I long ago learned long-form status reports were a pain to write; as a manager I discovered they were no fun to read either. Over time, I recognized that the problem was the collection of random data dumped into my blender, often shrouded in near-unintelligible business-speak. Creating a simple structure, where two of the three headings represented a bias for action, helped me turn the data into information – and saved a whole lot of time in the writing and reading of it. 

Data should underlie decisions wherever possible. But data can’t drive good decision making; for that, you need information you can understand, test, and act on.

(This column is modified from my forthcoming book, The Third Sigma.)