Big Data has some predictable challenges to overcome. Lone Star can help.

Nearly every day there is an “important announcement” of some NEW! analytics solution. Using AI (and perhaps magical pixie dust) this NEW! solution will let you monetize your data.

But, nearly every day, some important institution also announces the failure of their big data project, launched on the promise of a similar “important announcement” made a few months ago.

This pattern is consistent… but why? There are several reasons.

One reason is top-down pressure. Corporate boards want to demonstrate they have a “digital strategy” in place. It is the 21st century after all. This strategic playbook includes the idea of “Data as the new oil.” Believing you might be sitting on important oil reserves, you feel a need to start drilling.

But randomly drilling down into your data is even less likely to work than randomly drilling oil wells.

The truly strategic energy companies spend a great deal of time, money, and thought before the drill rig is ever set up. Are they expecting oil or gas? How will it be moved to market?

They are purposeful.

The Purpose is often overlooked in big data projects. Eager young data scientists produce what they feel are “interesting insights.” However, too often, they lack the operational context to know if they are being silly or impractical.

A great example is weather-driven operations. If you look for last year’s most highlighted locations for flight delays, shingle sales, and mobile power generation, you will find where the weather was very bad.

Though in reality, your database is probably not linked to NOAA or MET weather records. So, the eager young data scientist (EYDS for short) will tell you that Mobile, Alabama seems to be the magic market we should focus on. What the EYDS didn’t understand, however, is that this was a weather event, not a demographic or market result. Even if the data did link to hurricanes in 2017 and 2018, the EYDS might miss the fact that Mobile didn’t have hurricanes from 2012 to 2016 (see –

This is just one example of the silly things your eager, young data scientists will generate if you allow random drilling.

Another symptom is purposeless data collection based on the fear of missing out (FOMO).  More times than not, companies collect data for their own sake because they feel there may be something important in that data to use later… someday. FOMO is not purposeful. As the world changes, so does your data. If 99.9% of your old data won’t be useful in the future, you’re not collecting a data lake or even a data swamp. You have a data compost pile.

Data alone is not valuable. In fact, it may be just something expensive to store. Only analysis of that data can answer questions with high ROA. And that requires purpose. Whether your EYDS is your employee or works for a major consulting firm, it’s not their fault if you allow them to drill randomly into your old data compost pile. Ultimately, it’s your fault for letting this happen. Just as everyone needs a purpose to make life meaningful, your EYDS is no different. Please, make their lives better: give them purpose.

How do we find purpose in our big data projects?

One place to start is the “Return on Analytics.” At Lone Star, we start this hunt by asking about “Big Questions.” What Big Questions do you wish someone would answer? It doesn’t matter if the Big Questions are perfect, as they provide direction and increase the chances we are drilling in the right area. Our purpose is to drill down those Big Questions in your Big Data, providing customers with the Big Information they’re looking for.

An oil company might be looking for gas and hit oil.  They were in the right neighborhood for finding hydrocarbons, and their purpose had a high-value return.

Lone Star’s methods and software solutions help define the data needed and the right analytics to address purposeful goals. We’d love to help your organization deliver large Returns on Analytics.