Most companies, outside of large analytics-oriented shops like Google and Amazon, have struggled to find ways to translate Big Data analytics into bottom-line revenue. Part of the disconnect in creating bottom-line results from Big Data may be due to a lack of leadership sophistication around the concept of Big Data and its associated possibilities.

In a previously published article, Is Big Data Important for Non-Technical People?, I highlighted the importance of future leaders (even non-technical leaders) understanding the caveats of Big Data. I have received many requests for a layman’s explanation of Big Data to help support that assertion.

Forbes writers, as well as other subject matter experts, have used Lego blocks as an allegorical muse for Big Data – both because of the combinatorial possibilities of the blocks, and the ease of understanding. Herein, I present a simple explanation of Big Data utilizing a similar mental visual aid, to which almost everyone can relate, but I extend the metaphor in another direction.

Imagine someone dumps a large bucket full of assorted Lego blocks on the state of Texas. There are enough blocks to cover the entire state with numerous layers. Organizational leaders realize that these Lego blocks hold the key to strategic success. Accordingly, an employee receives the unenviable assignment of “sorting” the Legos. The problem is that the employee is never told how the blocks should be sorted. The employee simply receives the instructions to sort the Legos and provide a report describing what she finds.

With such a generic specification, the employee has any number of potential routes to accomplish the sorting task. Lego blocks have different colors. The employee may choose to sort them that way. Legos also have dimensions. Some are two prongs wide and four prongs long. Others have any variety of lengths and widths. Legos also have different levels of thickness.

Beyond physical attributes, Lego blocks have numerous potential esoteric properties. For example, some Lego blocks are more commonly used to build castles, whereas some are most often used for vehicles. Some are more likely to be choking hazards. Going one step further, each Lego block could have other unique attributes, such as having appeared in a major motion picture.

In the Texas-based example, each of the blocks also has a location. Some blocks were dumped in Austin and some in Dallas. Some blocks fell into cattle ranches. Also, since there are multiple layers, some blocks are on the bottom layer and some on the top, even within the same location.

All of these varied attributes together compose the “meta-data” associated with each block. It should be pointed out that meta-data are not always clearly defined or discrete. In fact, there are an infinite number of meta-data points (and combinations of meta-data points, such as green blocks in Austin) that could be documented for every piece.

Accordingly, the employee given the task to “sort the Lego blocks” could suffer from an immediate panic attack. Imagine further that not just Lego blocks are being dumped, but also action figures, lawn mowers, and best-selling novels – all of which must be sorted. Additionally, the dumping doesn’t happen just once – it is repeated every two seconds ad infinitum. Further, the dumping isn’t just happening in Texas, but across the entire world.

Note that the number of meta-data points also increases (although already infinite) as time passes and new dumps are added. For example, was a particular Lego piece part of the 2:00 PM or 4:00 PM dump? Did this Lego piece’s position change when the new dumps were added? Could the location of future lawn mowers be predicted by the current disbursement of Legos? In the end, without clear direction, the sorting employee’s mind could explode from the possibilities!

Therein exists the challenge for leaders; Big Data (the meta-data associated with Lego blocks in our example) are characterized by three major attributes: volume, velocity, and variety. The task of “sorting the Lego pieces” cannot, and should not, be given without further specifications. If it is, the data-sorting employees will provide something that may be interesting, but may not be useful for strategic purposes.

As a leader, you must be as specific as possible to ensure strategic measurement objectives are properly achieved. Telling an employee to “sort the Legos” will not be as valuable to your organization as telling an employee that you want to know about green blocks with specific attributes dumped at 2:00 PM in Austin, and how they predict the characteristics of future lawn mower dumps. This does not require an intimate knowledge of data-mining, but it does require knowledge of how Big Data variables and analysis integrate with your organizational strategy.


You may view the original version of this article at