CIO 100 Winners Turn Analytics Into Money-Making New Products
Mo Data stashed this in Data Valuation
This year's CIO 100 honorees collectively spent more than $502 million on their technology projects, and many of the winning efforts focus on using advanced analytics to create new sources of revenue, improve customer experience and increase competitive advantage.
Electronics manufacturer Celestica, for example, developed a supply chain analytics solution so beneficial, it has created a new line of business to sell the tool to customers. GE Capital rolled out a system that enables its midsize financing customers to manage their fleet costs with predictive maintenance and fuel optimization capabilities. Procter & Gamble built mobile analytics apps that worked so well for its own sales force, it now provides the technology to its distributors and other selling partners.
"A key driver is the increasing amount of data being generated each day," says Filippo Passerini, group president of global business services and CIO at P&G. "With the investments that have been made to measure business processes and produce the data, comes the expectation that analytics will drive actions that fuel business improvements," Passerini says. "Analytics allow us to turn this data into meaningful business insights so we can make better decisions."
Anyone who wants to keep their finger on the pulse of how a company is operating is investing in advanced analytics," says Celestica CIO Mary Gendron.
But these CIO 100 award winners are doing more than simply tracking business performance; they're transforming it. And it's not surprising that they're using innovative analytics to do it. According to a recent survey by business consultancy KPMG, 51 percent of technology executives say data and analytics will be the top driver of revenue growth in the future, ahead of cloud and mobile technology.
Analytics-driven business change will affect every industry, says Thomas Davenport, professor of IT and management at Babson College and author of Big Data at Work. "For 30 or 40 years, we have focused on getting transaction systems in place. Now we've won that war, and the new war is what to do with all of the data we have accumulated," says Davenport. "How to make better decisions on it and how to create new products and services based on it.
And our CIO 100 honorees are out in front spearheading the transformation. They're learning how to successfully deploy advanced analytics and racking up early success stories. In so doing, these CIOs and their organizations embody the ethos of today's strategic IT organization: putting themselves out in front, not toiling away in some back room in support of the business. Not simply coming up with solutions to business problems but helping the company make money, enter new markets and surpass the competition.
Starting at the End
It's tempting to jump right to analytics tools. But while the technology is critical, CIO 100 winners say it must take a backseat to the business drivers.
"You have to understand the business needs and then determine where technology can help," says Passerini. As of this spring, P&G had given 5,000 sales professionals a suite of mobile apps that offer instant, on-the-go visibility into data like in-store sales trends for Bounty paper towels or Bounce dryer sheets.
The program has led to $8.2 million more in sales and $6 million in savings so far, and is a hit among P&G's partners. In the past, sales folks had to drag around binders of printed--and dated--material when visiting customers. Sales reps can now access all available data about that customer right before their store visit to come up with plans to boost sales of certain products or rethink the mix.
At headquarters, a marketing manager can instantly view the progress of a product launch based on a store analysis performed in the field by sales representatives. "The faster pace of business requires better tools to operate in real time," Passerini says. "By identifying a pain point, we were able to use technology to better serve the business--and that's what we exist to do."
It's critical to have a clear target in mind for analytics, says Davenport. "Even once you have narrowed the field of application down to something like marketing, you still have to prioritize the choices," he says. They could include segmenting customers, targeting offers, avoiding attrition, optimizing the marketing budget, and so on. "You can eventually get to all of these, but you can't do them all at once," he says.
Eurpac, a provider of consumer product distribution and merchandising services, introduced analytics software that helps its manufacturing and retail customers better price their own products. The Tarot analytics system correlates retail data with information such as social sentiment and weather patterns. Eurpac handles consumer products--as varied as L'Oreal mascara and Weber grills--as they make they journey from the factory to the shopping cart. That's 30,000 products sold by 500 manufacturers through 2,000 retail outlets.
IT identified an opportunity to help Eurpac's customers more profitably price their products based on predicted demand. Tarot takes empirical in-house data like point-of-sale information and compares it to data on forces that might influence sales--weather trends that could delay garden purchases or an in-store tasting that might boost bourbon sales--to predict buying activity. "Our business had an objective sense of what influenced demand and we translated that into something quantitative," says Eurpac CIO Mike Skinner.
At Atlanta Public Schools, former IT director Glenn Melendez believed that with the right model, he could prove that some educators were better than others, to "start looking at human capital management in a whole new way." (Melendez is now an assistant director of IT solutions delivery at Ernst and Young.)
The school system, which serves approximately 49,000 students, deployed a system for analyzing teacher effectiveness, including a dashboard that provides administrators with data to make decisions concerning placement, professional development and compensation.
The dashboard, which was launched in the 2011-2012 school year, predicts which teachers will have the greatest positive impact on student growth in order to focus on retention and identify teachers with lower performance who need help. The ultimate goal--correlating teacher effectiveness with student outcomes--will take time to achieve, but initial results are promising. Sixth-grade math scores, for example, are up 12 percent since implementing the system.
The final business outcome must be the focus from the start, says Celestica CIO Gendron. Celestica hired Gendron from the Nielsen Company in 2008, in part because it wanted the benefit of her analytics experience in reimagining Celestica's role in the market, she says. Using analytics, Celestica improved inventory turnover and workforce management in-house, andA began to sell the system to customers.
That's an entirely new business -- launched by IT -- that helps original equipment manufacturers better manage their supply chains. In 2012, that new business helped its first customer manage a $4 billion supply chain and earned several million dollars, says Gendron.
"We didn't try to get the perfect data first," she says. "That's where we've gotten tripped up as an IT industry, spending millions on enterprise data warehouses and data cleansing. You don't start with the fuel, you start with the outcome."
That makes it easier to figure out what data is required. "We gained a lot of enthusiasm that way," Gendron explains. "That's how analytics came to life."
Partnering for Analytic Skills
Not surprisingly, the technology and domain expertise required to deliver successful analytics tools are in high demand, and CIO 100 winners are employing a number of tactics to acquire the skills needed.
Analytics is "a cross between art and science," says Skinner of Eurpac. The artists are the Eurpac employees in marketing or out in the field. The scientists are those in IT. Skinner created an analytics development team that blends both. Everyone was excited to be involved in the leading edge work, he says. "The difficulty was not in motivation. The difficulty was finding a way to get input from all the experienced people because everyone has a full-time job anyway." Skinner and his team worked around that by making it as easy as possible to participate in ways that did not interfere with everyone's day-to-day responsibilities.
For Melendez, formerly of Atlanta Public Schools, it wasn't that the right skills were hard to find. It was that he couldn't afford them on a public-school budget. Benefits like spring break and summers off only go so far. So he chose instead to train his existing staff in areas like Microsoft SQL Server Analysis Services. That, he says, "quite honestly, wasn't a bad thing. You want your resources to have the opportunity to learn new things, especially in IT."
Celestica had done very little analytics when Gendron came aboard, doing "rudimentary reporting at best," she says, which meant there was less focus on envisioning what might be possible with the company's existing data.
But that blank slate was also a benefit. What Celestica did have was an enterprise data warehouse that had hardly been used. "The company had been collecting data for 10 years," she says. "We were sitting on this treasure trove of insight." Gendron decided to let business intelligence vendors, including IBM, MicroStrategy and SAP, have a go at it, inviting them to participate in a contest to see what they could do with the data that could transform the company--and its customers.
Fail Fast, Learn Early
GE Capital Americas "has always been a data-centric company," says Kelly Shen, who is CIO of business intelligence there. "What's exciting now is that we have the opportunity to offer that to customers."A
Shen's IT directors and other business leaders sat down with the company's midsize customers who finance their fleets with GE to find out about their pain points. Their biggest challenge was cost management. Thus, Fleet Optimizer was born. The iPad app and desktop version combine analytics and geospatial visualization to help companies make better use of their vehicles. Based on data like make, model, year and mileage, the app makes recommendations about preventive maintenance and fuel management.
"We had the domain expertise and technology to make it happen and make it happen quickly," Shen says.
They got it done quickly by taking a iterative approach, using agile development processes for the first time. "Advanced analytics is best suited for this because you're dealing with a level of ambiguity," says Shen. "If you go through a traditional software development cycle, you see the data for the first time after you build something. With agile, you can start seeing things as you develop the prototype--usually days in."
The IT group delivered the first generation of Fleet Optimizer in 72 days. "The only way to know the value of [an] analytics [system] is to get it in front of end users as fast as possible," she says. "Fail fast, learn early, change strategy when it's not working."
At P&G, taking an incremental approach enabled IT to change course to meet business needs. "We have been providing technology to the sales force for more than 20 years, yet the explosion in mobile technology over the past few years gave us the opportunity to look at this space and approach it in a different way," says Passerini. "Among the biggest challenges was ensuring we could meet the broad business needs, yet making it easy to use."
IT started out thinking it would develop a single sales force automation app, but changed course to deliver an app ecosystem that could adapt to the different needs of each business. "This was also the first time we deployed tablet-based apps globally," he says. "We had to navigate technology availability across the world and create new processes."
Modular is the way to go when delivering analytics capabilities, agrees Celestica's Gendron. The first time the company offered its supply chain analytics to a big customer, it attempted to provide a full supply chain analytics suite. The customer was overwhelmed. So Celestica reworked the software so that customers could buy the system in components. "People need to learn to walk before they can run," she says.
Building Analytics Acceptance
For all the excitement around advanced analytics, the notion of machine-made predictions or management by algorithm can cause significant anxiety and reticence among users.
"You get a lot of skepticism and fear, especially in public education," says Melendez. That's one of the reasons Atlanta Public Schools used a $10 million grant from the Bill and Melinda Gates Foundation, rather than public funding, to finance a system that correlates student outcomes with teacher performance. It's also why Melendez and his team heavily involved teachers and administrators in the development of the system, which began with a prototype in 2009.
"This was not done in some back room with us emerging one day saying, 'Here's your new model. Best of luck!' That's not the way we did it," explains Melendez. "We took a couple of years to get input from our principals and educators."
A year of working sessions with district representatives led to a year of modeling and building a prototype that was used as a pilot. Then IT tweaked the model based on feedback. "There was lots of engagement and back-and-forth."
At Eurpac, users were embedded in the development team, and the analytics pricing software was developed in such a way that by the time the system was ready to roll, "they accepted it as their own work," says Skinner. "It wasn't something that was prescribed to them."
IT Steps Up
Starting with an end result in mind, taking an incremental approach and involving users intimately in development has enabled these CIO 100 winners to deliver significant benefits to their businesses from their analytics efforts.
At Eurpac, analytics success is defined by "how quickly we can generate superior insight so we can beat the competition or so when the end customer goes to the shelf, they reach for your brand over another," says Skinner. By those measures, it's working. But, he says, "the biggest benefits are beyond quantifiable, they're strategic." Eurpac has won new clients specifically as a result of its advanced analytics capabilities, including Jose Cuervo and Microsoft, which came to Eurpac from a competitor, he says.
Those kinds of business results have raised IT's profile at these winning companies, and ultimately resulted in new ways of working not just with the business, but as one with it. "The whole divide between IT and the business and external customers? We've knocked down those walls," says Shen.
The Fleet Optimizer is just the first in a line of analytics products the company plans to sell to customers. IT is developing similar products for original equipment manufacturers and marine industry dealers, helping them create the right product mix based on demographics or external conditions, like macroeconomic trends and weather. "We recognize the need for roles and responsibilities, but we're all collaborating," Shen says. "We're all interacting with customers at the advisory board meetings. We're not working behind the business."
Prior to IT's introduction of advanced supply chain analytics at Celestica, the organization was seen as an enabler of the business. With this project, "IT stepped up and became a leader," says Gendron. For example, the supply chain analytics business is staffed by IT professionals who sell and service the product. "By bringing IT closer to the customers, we're not only running a more effective internal IT organization, but changing the company's role in the marketplace."
"That's the great lesson from online companies in Silicon Valley like Google and LinkedIn--you can take all this data, analyze it, and make it into products and services that you can either sell to customers or use them to make your other offerings more attractive," says Davenport.
At Atlanta Public Schools, teacher tenure--always a hotly debated issue--used to be based on years served. But, Melendez says, "number of years did not necessarily correlate to teacher effectiveness. Now we can look at student outcomes to earn tenure." Administrators can now create educator teams, pairing up strong teachers with weaker ones and elevating the overall team.
These award-winning analytics applications are just a first step in harnessing data to improve business. "As you start to roll this out, you can start to dream about next steps. You can start to look at all kinds of correlations," says Melendez. "I don't know that you ever get to a point where you're finished."
The application of analytics to transform the business is an evolution, not a revolution, says Gendron: "It's an ever-changing world and you've got to keep your eyes wide open."