"The CIO can say, 'Give me your cash-flow data and I'll mix and match it with my data,' and it starts to build a relationship," Sahoo says. IT can also build dashboards of data supplied by different business units to help managers see the value of combining and analyzing data that currently may be squirreled away in a departmental data silo.
In this respect, IT is integral to building both credibility and trust in data analytics, says Thomas.
"Decision-making always comes back to a single source of the truth. Otherwise, teams spend a lot of time on data skepticism and discussing why data is invalid," he says. "When you put all the data together, looking at a number on its own becomes irrelevant."
At Intel, which is widely regarded as one of the most data-driven enterprises in business today, "we see big data and analytics as complementing better and faster decision-making," says Ajay Chandramouly, whose job title is big data evangelist.
"At Intel, it is in our DNA to be data-driven. We don't see two different camps -- the HiPPOs and the quants. We see that as a false choice," Chandramouly says. "What we try to do instead is use the insights from advanced analytics to ask the right questions. When you frame it in that perspective, it tends to minimize conflict."
5. Communicate your successes, but keep fine-tuning.
At CNA, Wolfe publicizes his unit's ever-rising fraud detection rate, to both employees and customers, as a means of reinforcing the company's focus on hard data and analytics.
"I have committed to having every member of the special investigations unit get out in front of a customer to talk about our program," he says. "We're a commercial carrier and many of our insured want to know what we're doing to protect their bottom line, especially regarding workers' compensation claims," he says. "Customers love to hear what we're doing with new technology and what is going on behind the scenes. The technology has given us a competitive edge in many [customer] meetings because not all companies are doing this."
At USAA, Mowen says a common practice is to communicate the results of successful projects carried out by the chief data and analytics office beyond the project's stakeholders.
"The more people who understand that we ran tests on data, the better informed the whole organization becomes," Mowen says.
Another fine-tuning technique: "When we really run into strong HiPPOs, we create experiments or tests where we can test their hypotheses. We have a specific test-and-learn team in the CDA that does this," he says.
The bottom line: More and more, "people state their source clearly on all documentation," Mowen says. "They note that the facts came from the CDA."
Julia King is a Computerworld national correspondent.