• December 15, 2016

Act Fast: Turn Data into Action Without the Lag

Holding out for perfect data is the enemy of good outcomes. 

What if once a building’s blueprint was complete and the building materials were purchased, no one bothered to figure out how to use the crane or power tools and actually build it? It’d be a waste of potential, not to mention time and money. It’s the same for enterprise data.

Data and analytics can bring unprecedented enterprise value, yet many organizations fail to use insights generated from data analytics to drive action. According to Forrester Research, just 29 percent of data architects say their organizations have solid strategies in place to create business value from analytics.

Turning data into action can be cumbersome. Too many organizations think they need perfect data to gain usable insights from their analytics processes. Yet often even limited data sets can yield positive predictive results. To paraphrase Voltaire, holding out for “perfect” data is the enemy of good business outcomes.

For example, Telstra Corporation Ltd., Australia’s largest telecommunications and media company, harnessed analytics in its call center to boost customer satisfaction levels. It started by analyzing unstructured data sets containing customer call audio files. Through this process, the company was able to quickly detect early signs of trouble by analyzing patterns associated with customer tone, emotion, and word choice. The analytics strategy enabled Telstra to address problems before they led to disconnection requests, accelerate issue resolution, and reduce customer complaints by 24 percent. Bonus: Revenues jumped 11 percent since deployment.

Dodging Traps

But while analytics strategies hold much promise, they also harbor pitfalls. Many enterprises fall victim to conflating analytics with technology. Effective analytics systems are not driven by good data sets and specific applications and algorithms alone. Successful strategies cannot be achieved simply by implementing specific technologies. You have to lay solid foundations.

First, identify the processes you want to improve and what outcomes you expect to achieve. What data is needed to drive these improvements? What strategies can you deploy to effectively convert data into predictive insights? How can you effectively spur action from these insights?

Typically, the insight-to-action process happens in stages. The enterprise crunches and analyzes data. Then it presents the analysis and discusses the results. From there, it explores actions. Once the most effective actions are determined, they are delegated and executed. In the final stage, the enterprise analyzes the resultant outcomes.

Getting Caught in a Time Lag

The problem with this scenario is that there are too many handoffs and individuals and teams sandwiched between insight and action. The process creates a substantial time lag. Every moment that separates data insights from action lessens its impact and narrows potential opportunities. As machine learning drives improvements in data analytics processes, many of these “human intervention” steps will be automated. Ideally, these systems will be capable of delivering actionable insights directly to those who are best suited to take the actions that will propel business outcomes.

In the analytics game, speed is essential. Success demands a rigorous focus on the disciplines and technologies that drive insights, and the operational processes that turn those insights into executable actions. With a well-engineered analytics process in place, the enterprise can successfully collect and leverage its data to improve productivity, rein in costs, drive innovation, and boost revenue.

Like this story? Read about customer insight from predictive patterns.