It’s 2014, and Big Data analysis is still big news. Businesses in every market segment are scrambling to use this insight to create “right-time” experiences for their customers and internal staff. When done well, Big Data analysis gives companies a clear competitive advantage.
However, for many organizations, Big Data efforts have gone awry, transforming the Big Data buzz into a buzz kill; derailing initiatives before they even get off of the ground.
According to recent studies, more than half of Big Data analysis projects never get completed, with many others falling short of their objectives. All that’s left of these efforts are corporate leaders shaking their collective heads, wondering where things went so wrong.
So, why do so many analytics efforts fail? What’s the best way to capitalize on the Big Data opportunity while navigating the potential roadblocks that could slow progress? We spoke to some industry experts to find out the answers. Their insight could save you some time, cost and heartache as you embark in your own Big Data journey.
Identify Your Goals Before You Jump In
Our experts all agreed: One of the biggest causes of Big Data disasters is jumping in too soon without clearly defining what it is you want to accomplish.
“In a typical scenario, the CEO pings the CIO and says, ‘we need to be doing something with Big Data.’ The CIO runs out, buys the technology tools and delivers analyses that have little to no value to the business,” explained Anand Sahay, head of IT Services for IGT. “Big data analysis cannot be an initiative of the CIO. It has to be a business initiative executed by the CIO, based on a well-defined business case.”
And we’re not talking “gain customer insight,” “increase productivity” or “reduce operating costs.”
“While many companies have the big themes right, to be successful, they have to connect those big themes with the granular idea of how to add value to the business,” explained Vikram Duvvoori, corporate vice president for HCL America. “If you want to increase productivity, define where—in what areas. Gain customer insight to accomplish what? Companies often get caught up in the process of collecting and analyzing data, instead of focusing on how that insight can help the company do business in a different, better way.”
Although creating a strong business case is fundamental, it’s only step one of an effective analytic process.
Block and Tackle First
“Jumping in to analyze large volumes of diverse data without first having a true business use case and some level of data governance in place – what I often refer to as part of the basic blocking and tackling – is a recipe for disaster,” explained Scott Schlesinger, senior vice president and head of Business Information Management for Capgemini. “Once you establish your business case, you have to identify what data you need to grab, how you’re going to bring that data into the organization, what tools are best suited to capturing the type of data you need, and how are you going to store all of this data.”
One of the biggest obstacles is the way companies currently manage their data.
Previously, companies organized data in siloes. Now, they not only need to manage that data but also weblogs and feeds from Facebook, Twitter and other social media sources,” said Guru Shashikumar, director of product management for Oracle Managed Cloud Services. “For companies to take advantage of the Big Data opportunity, they have to tear down the walls—both in the way data is organized and in the way insight is shared.”
They also need to put strong governance processes in place; a critical factor too many organizations currently ignore.
“So many companies buy a solution but don’t have a data governance process in place or the expertise to set it up,” Schlesinger of Capgemini said. “You have to create a strong control framework so you know that the data you’re analyzing is good; that your models are sound, and that your processes adhere to privacy standards, which vary from country to country. Big Data analytics is such a hot topic that it’s easy to get caught up in the technology, but without the process, program and the right people, you’re going to spend a lot of money and not get the outcome you want.”
If you’re new to Big Data Analytics, don’t try to go it alone. The right partner can help you create an analytics playbook, set up a formal governance program and train your people on how to manage that program going forward.
Integrate, Analyze and Pay Attention to Context
Integration of data, both in collection and how insight is dispersed among your various communication channels is essential—lest you irritate the very customers you’re trying to delight.
“Imagine going to a specific resort and having a fabulous time. A week later, you get an offer from that same resort for the same vacation you booked for $5,000 for $4,200. You’re going to be angry and probably demand an $800 refund,” Schlesinger of Capgemini said. “All of that does more to damage the relationship and eradicate what should have been a good customer experience.”
Context is also critical.
“Creating a right-time experience is more than customer relationship management. By applying contextual information—what is the customer doing right now, what does he or she need, what is his or her attitude at this moment—companies can truly transform how they interact with their customers, prospects and employees,” Shashikumar of Oracle said.
A handful of data sources, and one or two internal departments using the insight, could lead to disaster.
“You don’t want to send an offer to buy jewelry for your wife to someone who has recently changed his Facebook status from married to single,” Sahay of IGT said.” At the same time, you don’t want a customer to get multiple offers or communications from different departments in isolation. If a passenger is stranded at the airport, you want to send an apology, with up-to-date flight information, and also send a discount coupon for a massage at the airport spa while he or she waits. That coordination between departments is a process issue that has to be put in place before an organization embarks on a Big Data program.”
Change Your Corporate Mindset
But even if you create the perfect analytic structure, if your company isn’t positioned to respond to the insight, you won’t see the benefits.
“Companies have to embrace a data-driven culture, in which employees are empowered to make decisions quickly based on the analyses they receive,” Duvvoori of HCL America said. “If you’re an organization that’s used to spending hours in meetings discussing logo color, this is going to be a cultural shift. Instead, you’ll test that logo color or ad or product idea and react to the response.”
Instead of that end-of-the-month dashboard with concrete facts, the power of Big Data analytics is in the experimentation.
“With Big Data analytics, you really don’t know in advance what questions to ask and what information requirements to give to your IT. It’s more about discovery and exploration; about understanding dynamically changing customer behavior; spotting trends—and responding to those trends,” explained Dragan Rakovich, CTO, HP Information Management and Analytics Services. “The analysis is not the end point. Companies have to have processes in place to continually monitor and respond to what’s happening. That’s when they’ll gain the competitive advantage.”
Leverage Your Outsourcing Partners’ Expertise
If you’re don’t have analytics expertise, work with a trusted, experienced partner who does.
“Customers should discover the true business value in their own data before making capital investments into hardware and software infrastructure. To achieve that, they should look for their partners to provide them with access to integrated Big Data analytics platforms and data scientists to test drive their Big Data use cases, ” Rakovich of HP said. “This approach allows customers to jump-start their innovation in Big Data analytics, and ultimately create a solution that can respond to continuously changing information needs. “
At the same time, if you are currently outsourcing specific functions, make sure your provider is using Big Data analysis to continually improve performance, mitigate issues and reduce costs.
“In today’s world, productivity is the chief currency, and by applying Big Data analytics, we can enable our clients to be more productive and make better choices,” said Paul Gleeson, vice president of Global Strategic Sales and Operations for Unisys Global Managed Services. “As an organization, we collect an enormous amount of valuable data. Using this insight, we can tell our clients which software or hardware is problematic and which is robust. We can also use Big Data to drive problem resolution down to the SKU level and hardware manufacturer.”
According to Gleeson, outsourcing providers can add real value to clients by using data in a whole new way.
“For one client, analytics showed us that contractors were 1.6 times more likely to contact the call center than employees. After digging deeper, we discovered that the cause was actually training. Onboarded employees had two weeks of training, whereas contractors had none,” Gleeson said. “By adding a process for contractor training, the company not only saved money in help desk support but also improved productivity.”
By collecting and analyzing data from a wide range of clients, service providers can identify universal patterns and more quickly introduce process improvements that enhance productivity and reduce costs.
Start Small
So, what is the best way to turn the Big Data challenge into a Big Data opportunity?
Our experts’ advice: start small and start right.
“Everyone needs to take a deep breath. Pick an objective, get the right people involved and test multiple times before a big roll out,” Schlesinger of Capgemini said.
“Instead of making a huge technology investment, consider open source tools, or hosted SaaS tools, where you pay by transaction,” Sahay of IGT said.
“Clearly define which business problems you could solve with data insight,” Duvvoori of HCL America said.
“Get top down support, and involve the right people throughout your organization. Clearly define roles and responsibilities as well as a SWAT team of business and technology personnel to drive the effort,” Shashikumar of Oracle said.
“Use your initial success to build a data-driven culture,” Rakovich of HP said. “Again, before making that initial investment, before creating the business case, try out the technology and test your data to prove your business use case.”
“Make sure you measure your results and continually refine your processes,” Gleeson of Unisys said.
All great insights–coming just at the right time. Now, what are you going to do with the information?
