Why one CIO hopes self-service analytics will make IT irrelevant

Merck Animal Health is using cloud analytics software to gain insights into its supply chain, including such crucial data as whether its customers received their orders in a timely and complete fashion. Eventually, the company's IT department hopes to enable business users to manipulate the data themselves, says Dave Williams, CIO of the $3.6 billion Merck unit that makes animal vaccines and other pharmaceutical products. Until then, Williams will have to continue weeding out data quality errors and helping manage the IT ownership changes associated with adopting cloud software.

"When we build self-maintained analytics, the business users can do it without IT's support," says Williams. "We should strive for that so that we're irrelevant."

By “irrelevant” Williams means reaching that hallowed state of IT nirvana in which hardware and software run in the cloud, tended to by several vendors, with little care and feeding by the IT department. For the foreseeable future, though, Williams must contend with the kind of data quality issues that many CIOs are struggling with as they increasingly look to software to find morsels of useful data. Some 41 percent of 550 executives surveyed by the Economist Intelligence Unit said they struggled to maintain data quality. Thirty-three percent suggested that managing the data alone was a challenge.

Economist Intelligence Unit

The joys of implementation

Williams, who in a previous role at Merck spent nearly 10 years building up analytics, knew how difficult it was to get business data out of the company's business software and into data analytics software from another vendor. The data must be normalized, ensuring that the business language aligns on an apples-to-apples basis, and scrubbed clean of erroneous information before it is moved.

[ Related: SAP unveils SaaS analytics platform ]

As part of a plan to “triple down” on analytics, Williams began using FusionOps to monitor the efficiency of a crucial supply chain that ferries products to distributors who sell them to veterinarians, farmers and pet owners. Since implementing the tool more than a year ago, the unit has improved its ability to distribute products on time and in full, to track payments, and to detect slow-moving products. It uses the software to measure accuracy for the last several months of orders and align production with market demand. Plant level managers use the software to track shipping and model schedules for upcoming weeks.

Although he says he does not yet have metrics to demonstrate supply chain improvements as a result of using FusionOps, Williams says anecdotal evidence from business colleagues suggests that FusionOps has “helped us move the needle” for measuring “how customers are experiencing us.”

[ Related: Partnership powers predictive and prescriptive analytics ]

It’s also saving time in the IT department. Traditionally, Williams’ team would figure out where data was in their SAP ERP system and how to extract it, normalize it and analyze it. Then they would configure servers and business intelligence applications to build reports from the data. FusionOps’ SAP connectors ingest data into the vendor’s cloud in a standard, normalized view based on pre-defined business rules. Then algorithms crunch the data for insights and populate the findings in bar charts and other graphs in Web dashboards. “The amount of time we spend analyzing versus hunting and gathering data has shifted dramatically,” Williams says. “Now it's ‘how do we improve this metric to drive value on the balance sheet.’”

Combatting change management with collaboration

Initially, Williams’ introduction of FusionOps put off IT staff accustomed to setting up its own systems and combing through the data. “Cloud makes people uncomfortable,” Williams says. “People on the tech team see it as threat.” Moreover, changing how the department delivered information presented a challenge for business users who were accustomed to viewing data a certain way and become frustrated when they can’t find data because a different tool has a different user interface and format.

Williams says he addressed this issue by partnering with the business and his team, getting senior leaders from manufacturing and other departments behind the effort before it started, and easing the rank-and-file users into the transition. It helped that Clark Goldestani, Merck’s global CIO, recommended the product. “We had top-down sponsorship right out of the gate,” he says.

And while the IT department is still adjusting to the change, employees in manufacturing, marketing and procurement love the speed and agility with which FusionOps enables them to generate business insights. In fact, they love it so much that they are putting more pressure on the IT department to add new features and functionality. Williams is prioritizing the requests.

Williams says he’d like FusionOps to add functionality that will enable, for example, manufacturing personnel to query deep into SAP without help from IT. He says that while FusionOps isn’t quite prepared for this capability, some recent demos he’s seen of future functionality left him confident that this type of self-service analytics will arrive at Merck Animal Health in the near future. That, he says, will help shift the IT department’s emphasis away from building things internally and enable them to focus on more strategic projects, ostensibly to drive more value for the business.