The Top 5 Most Misused Data Buzzwords

“I’ll be honest. When my CTO came in a few years ago and started talking about a ‘data lake’, I thought he was kidding.” This is Jeff, a senior executive who caught up with us recently to talk about some of the challenges of staying on top of the latest tech developments for business. “I mean, first we’re saving everything in a cloud and now we have data lakes? What’s next - file fields? Data hurricanes?”

The tech industry can be ruthless when it comes to the latest jargon. In fact, one of the most common challenges we hear from Program Managers, Directors, and CEOs alike are that they are expected to stay up-to-date in their own field while somehow managing to keep tabs on tech enough to be able to adopt the latest advancements for business. Jargon can be a major barrier for some. Sometimes it’s easier to grab for the latest buzzword and check the facts later. This can lead to serious miscommunications at times, though. Did you really mean to propose applying machine learning to a problem or just analytics? Today, we’re tackling the top five most misused buzzwords to help you through your next meeting with your CTO.

Predictive Analytics

We’ll start with one of the most common phrases you’ll hear when it comes to data: predictive analytics. On the surface, this one seems pretty straightforward. Through predictive modeling and forecasting, this practice allows analysts to interpret available data to arrive at a prediction of future behavior. For example, businesses can predict retail purchases based on past behavior; analysts can assist doctors or insurance companies in predicting future health events in patients’ lives based on their medical history and the histories of other similar patients; entertainment companies can develop content that readers are likely to consume based on their previous reading habits and preferences. The catch is that this only applies to situations where you have data to begin with! We see this term thrown around most often by businesses getting ready to launch a new product and confronted with the question: what sales can we expect in the first quarter or first year? “Let’s just get a few folks in here and get started on predictive analytics” may sound great, but won’t actually be possible if you don’t have any historical data to get started.?

Artificial Intelligence (AI)

We’ve noticed a growing trend in professionals tossing around artificial intelligence almost as a synonym for problem solving or even allowing computers to produce an independent analysis of data. AI is actually a process that is directly the result of user input and training. Artificial intelligence is created by combining large amounts of data with fast processing and algorithms that allow the program to learn automatically from patterns or features in data. Artificial intelligence is what gives computing power to predictive analytics and has many applications. AI can generate reports and simply data so that it is easily understood by more than just data scientists. All of this is made possible by machine learning algorithms, which brings us to our next buzzword.?

Machine Learning (ML)

Believe it or not, the most common misconceptions about this term are that it 1) is just a synonym for artificial intelligence or 2) possibly applies to machinery. Machine learning instead describes the process of teaching computers how to learn by interpreting data, classifying it, and making adjustments based on successes and failures. One popular example of this is voice recognition software. Every time you ask Siri a question, there is a process at work that translates auditory input (your voice) into data, interprets the data, and identifies an appropriate response to the question. In order to “teach” the computer to be able to process this auditory input, programmers had to provide it with thousands of samples of recordings and map that auditory information onto data. To put it simply, machine learning operates on previous programming while also adjusting to new conditions or changes.

Data Lake

The next buzzword is arguably the most ambiguous if you’re not a data scientist. As cool as a Matrix-looking body of water would be, a data lake actually deals with storing data in its native format for exploration, offering an unrefined view of data. To the highly trained eyes of data scientists, a data lake offers solutions to potential problems. Each data element in a lake is assigned an identifier and tagged, and when a potential problem or question arises, the data lake can be used for relevant data and that data can be analyzed for a solution. So, how is a data lake different from a data warehouse? Well, let’s imagine for a moment that you are collecting and storing wood that is left over from construction projects. If you are following the “data warehouse model”, you might already have storage shelves in place in your warehouse with compartments that are a set size (let’s say 6” x 6” containers lined up on each shelf). Your supplier drops off the leftover wood that is in a variety of sizes. When it arrives, you have to cut it down into 6” x 6” pieces so that they can be carefully stored in the containers. If you are following the “data lake” model, though, then you would just accept the wood of varying sizes as is and hold on to them so that they can be used for different types of projects. To use a more technical description: in a data warehouse, the data structure and schema are defined in advance and all of the data is cleansed and enriched; in a data lake, the structure and schema are not defined in advance and a variety of data analytics can be performed on the data.?

Data Storytelling

Data storytelling is the practice of creating a narrative around data and providing additional context to influence how it is understood. Charts and graphs are excellent for presenting information visually, but they do not support any particular course of action. Imagine that you are presenting information about last quarter’s sales to your board of directors. Your team gathers information on which items sold, how many, the total profits, manufacturing costs, and a synopsis of customer feedback. You convert that data into a graph and show it to the board. Just showing a chart is not enough to tell the story of your company’s performance or motivate the board to take any particular course of action. Instead, it would be helpful to explain: were sales higher or lower than the previous quarter, is customer feedback positive, which products are popular and which need additional marketing support, who is the average consumer, etc.?

The fact is that leaders in most industries are now expected to be comfortable discussing the latest tech advancements that are essential to stay competitive in the modern world. Are you looking for more tailored digital literacy training for yourself or your team? Email [email protected] today to learn more!