In our Big Data world, we are awash in data. We have a lot of it. It comes at us quickly. And it comes in different forms. Those three Vs (i.e., Volume, Velocity and Variety) represent significant hurdles to businesses in their race to extract value from their data. Each of those hurdles have been removed (lowered?) with new advancements in technology and science. With parallel processing capabilities, machine and deep learning algorithms and edge computing, those 3 Vs should really be written in lowercase.
The Meaning of Your Data
Data are more than a string of numbers. They have meaning. They represent something of interest; they might represent predictors of events or outcomes that are important to your business. Numbers could reflect sales performance. They could represent customer satisfaction. They could even reflect healthcare quality. Ensuring your data represents something meaningful to your business is essential to finding value from your data.
The meaning of numbers is about the veracity of your data. Every time we use a metric, we need to ask, “What does that number mean? What does it measure?” For example, we could arrive at a measure of “customer satisfaction” in different ways. Customer satisfaction could be gleaned from ratings from customer surveys, culled from content of Twitter feeds or extracted from transcripts from call center conversations, to name a few. These measures of customer satisfaction are vastly different from each other. Survey ratings are purposeful measures designed to reflect general satisfaction while sentiment scores from Twitter feeds are based on data we have little control over and reflect fleeting levels of satisfaction. While these metrics all fall under the umbrella of “customer satisfaction,” they mean different things.
Subject Matter Expertise
Technology alone will not solve the veracity problem. This is where subject matter expertise plays a crucial role in helping you understand the meaning of the metrics. You need subject matter experts to help evaluate your measures. These experts, with their deep knowledge in a particular field, understand nuances about different sources of data and types of metrics and what they mean to solving a specific problem. It’s no wonder that subject matter expertise is one of the three pillars of data science.
Don’t forget that a number (metric) is more than just a number; it represents something of interest to your company. Be clear about your metrics and what they mean to your business. Appreciating that numbers have some underlying meaning is the start of understanding why the veracity of your data is so important. You can read more about veracity here.
