4 Ways to Create a Competitive Data Strategy in 2020
We have entered the age of analytics.
In this age, the businesses that utilize their data to make better, faster decisions are the ones that will succeed. For example, tech-savvy financial institutions are analyzing their data to see not only where their customers are spending money, but how much and on what. They know before a credit score drops that a direct deposit was missed, and they can track when business deposits have increased and cross-sell accordingly. This enables them to offer their customers the right products and services at the right time, creating more value, cultivating stronger relationships and beating the competition in the process.
No matter what industry you’re in, a solid data strategy is a necessity. I recently sat down with Dr. Taylor Nadauld, Associate Professor of Finance at Brigham Young University and nCino’s Chief Economist, who offered some guidelines for those organizations just beginning to dig into their data.
1. Save Everything, Just In Case
“All data is valuable. But not all data is of equal value,” Dr. Nadauld says. This is why the first question business leaders usually ask their analytics teams is what kind of data they should store. The answer? As much as you can.
When it comes to analytics, the more data you have, the better. “You don’t know what you don’t know, what questions you’ll want to ask, or what you’ll need three years down the road,” Dr. Nadauld says. “But one thing is certain — at some point, you’ll say ‘I wish I had that data.’”
Just because you’re keeping every bit of data doesn’t mean it needs to be immediately accessible. Dr. Nadauld recommends storing the data you’re currently using in the short term memory of your operating environment, and keeping historical data tucked away in a storage facility until you need it.
2. Understand the Big Picture
“When our analytics team gets a new data set, the first thing we do is figure out the story that the data is telling us,” Dr. Nadauld says. This starts at a high level. Ask yourself what variables you have, what is the scope of your data, and what is the big picture? “For example, if we’ve got a bank’s data on loan applications, the first thing we’d look at is how many applications were received, how many were approved, and what attributes those approved loans share. This offers a high-level understanding of what’s going on.”
Summary and statistics can go a long way toward telling the story of your business. Once you understand the big picture and the broad strokes, you can begin to drill down into the finer details.
3. Begin with a Business Use Case, But Don’t End There
Analytics should be leveraged to serve a specific business case and solve a specific problem — not the other way around. “Data geeks love to experiment and tinker, and will start looking at stuff that may or may not be helpful to the business’s goals,” Dr. Nadauld says.
This means a project manager or executive should approach the analytics team with their business needs, rather than asking what the data can do. At the end of the day, your data must provide actionable insights that align with your mission, increase your business’s value, and serve your audience.
While it’s best to approach your analytics team with a plan, it’s also important to be open to new ideas. “Data can get us to ask questions we never even thought about asking,” Dr. Nadauld points out. “We think we know what we want to ask, but then we look at the data and the story changes.”
Have a goal, but be open to new information. As long as your team is aligned on its goals, the new ideas they propose might help further your mission.
4. Trust Your Intuition
While analytics can help you make data-driven decisions, there are some potential pitfalls to be aware of along the way. Most of us have heard the old adage, “Correlation does not equal causation.” Just because the data indicate that variables are correlated does not mean the variables have a cause-and-effect relationship. Before acting on any information, make sure you work through associations and correlations carefully so that you’re seeing the whole picture and not jumping to conclusions.
Another rule that Dr. Nadauld recommends following is that the most basic relationships in the data should be manifest in even the most complicated models. Therefore, if a model is giving a weird answer that seems off, check your model rather than your intuition. People generally have good intuition, especially when they’re experts in their field. We built computers and programs to help us solve complex problems, but don’t assume the AI is always getting the economics right. Let your intuition serve as a guardrail when following the track laid by complex models.
The use of analytics is changing our understanding of the most basic aspects of human behavior in ways that will drive business going forward. It’s time to harness your data and see where it leads.
