Data volumes in global enterprises are growing exponentially as workloads move to the cloud, IoT brings in big data through sensors, and other digital trends. In the past, we may have analyzed spreadsheets with 30 columns of data, but now, making sense of all the available data to inform new business initiatives has become daunting.
Here are five questions you might find yourself asking while on your journey to analytics maturity:
Does analytics mark the end of the gut feeling?
Though daunting, making sense of huge volumes of data is lucrative. Companies that use data to inform decisions significantly outperform peers. However, a good number of decision makers still do not take full advantage of data to drive business decisions and instead continue to rely on gut feeling and recent anecdotes. Experience-based intuition is definitely valuable, but undervaluing the power of analytics is courting disaster. Analytics provides a more complete answer than prevailing assumptions and business truisms.
For example, sales vice presidents may answer that the difference between winning or losing accounts hinges on having experienced sales reps. That assumption has an element of truth to it, so you are likely to find some data that correlates to it. But other factors also matter, including language, market, channel, industry and competitors. There are patterns uncovering why companies win or lose, and executives must be open to using tools that can pattern data for a complete answer to their business question.
Whose yard is analytics — IT’s or the business’s?
In the past, IT would manage large volumes of data and use it to serve up reports to the business. But over time, business users began saying, “We have expertise in our departments, just give us the data, and we’ll analyze the data ourselves.” As a result, friction can be created when IT tries to control the data but the business wants open access. A better alternative is what we can call the “shop for data” paradigm. Business users have access to governed data managed by IT, can easily join it with any outside data they choose, and then explore it for business insights. What’s important is it’s always clear what was the enterprise data and what was outside data, so when people make decisions based on the data, they always know where the data came from.
Is my enterprise’s data enough?
The signals that underlie business performance may not all be in enterprise data, but rather in market data like social media or even a combination of enterprise and market data. One example arises if a competitor has a large product recall. That event may not register in your internal data, but it will likely be noticeable in social media, where people express their dissatisfaction with the product and their intent to purchase another product. Weather data is another good example. If a big snowstorm is approaching and you’re a car rental business, you’ll want to combine current and historical weather data to anticipate customer behavior and the corresponding impact on fleet management.
Who will make more decisions — humans or machines?
People and machines handle 40% of strategic decisions today. This mix will vary on the basis of individual business needs. For example, machines have become quite accurate at identifying suspicious activity on credit cards. Fraud detection models are constantly being improved, thanks in part to advancements in machine learning. Models become smarter and smarter all the time, and they’re at the point where machines can act on their own.
But that won’t be true in other areas, such as healthcare. For example, doctors can work closely with machines when a patient presents with a rare disease. The machine can process all the relevant data worldwide — something that’s impossible for a person to do as quickly — and bring the results to the doctor. The doctor would combine that information with his or her own insights about the patient and devise an appropriate treatment.
Solve problems or look for patterns?
With traditional predictive analytics tools, business users typically came to the data science team with a specific question and requested a piece of analysis that presented an answer, such as use cases around fraud, customer churn or employee attrition. In other words, shining the spotlight on a business problem. That is beneficial, but what’s even more valuable with all the new data is turning on the stadium lights and illuminating your business to find everything that’s interesting and has a pattern. There may be areas where no one’s looking for insights about improving product quality or better segmenting and targeting customers. Companies must use new and smarter analytical tools to both narrow in and zoom out of the data to influence decisions.
Legal disclaimer: The views expressed are those of the author only and do not represent the views of any of the member firms of Ernst & Young Global Limited.