Over the years, I've written many posts about data, analytics and humans and how all three combine to deliver significant business insights. In this post, I’ll recap some key points I've made about each of these (along with key points from others, too). Along the way, I'll show how the three work together to form what I call the insight equation.
We love data. And we love putting words like master, big and meta before it – and adding words like management, quality and governance after it. We also love collecting a lot of it, especially in data lakes. Data lakes are storage repositories for holding big data in its native format, often before data structures and business requirements have been defined for its use.
Perspectives on big data quality vary. Some argue that big data has a smaller margin of error (i.e., less statistical error) while others argue that big data eliminates sampling bias. While there’s some merit to these arguments, neither can prevent a systematic bias from overwhelming big data. Which is why it’s important to remember that while a data lake is a container for multiple collections of varied data coexisting in one convenient location, you still need data management to generate the most business value from data. That includes data preparation, metadata management, data integration, data cleansing and data lake governance.
Analytics are amazing. Most enterprises employ multiple analytical models in their business intelligence applications and decision-making processes. These analytical models include descriptive analytics that help the organization understand what has happened and what is happening now, predictive analytics that determine the probability of what will happen next, and prescriptive analytics that focus on finding the best course of action for predicted future business scenarios.
Analytics illuminate business conditions, because data is useless if you don’t have a business context to interpret it. Analytics enable data-driven organizations to make better decisions and produce better outcomes.
People play a pivotal role in data-driven businesses – good results rely on more than just good data and analytics. Human oversight is needed before data and analytics can drive good decisions and actions. Human oversight includes:
- Data stewardship,
- Testing and validating analytical models.
- Maintaining data catalogs, business glossaries and dictionaries needed to track data lineage and facilitate collaboration.
Of course, it’s only human to be at least a little resistant to going where data drives us, especially during one of those many times when data contradicts our intuitions, superstitions, beliefs, stories or – most important – our biases. In fact, overcoming bias may be the most essential element of the human involvement in successful applications of data analytics.
The insight equation
Individually, data, analytics and humans are capable of great things. But when combined, they can generate comprehensive insights for your enterprise. I call it the insight equation:
Data + analytics + (humans - bias) = insightsRead more about data management