Between 2005 and 2007, Jim Braun was CFO of OmniTRAX, a privately held railroad, transportation, and real estate management company. OmniTRAX owns shortline railroads (less than $500 million in annual revenue), but a significant percentage of its income in those years came from leveraging the land the company owned around its tracks. If, for example, a manufacturer shipped its product on OmniTRAX rails, it made sense for that manufacturer to locate a warehouse (or even a factory) near the tracks, and lease (or buy) the land (or building) from OmniTRAX to do so.

That model, in which OmniTRAX used real estate for top-line revenue generation, “worked out great,” Braun says, and the company was investing heavily.

Of course, in 2007 the real estate market was on the brink of an historic collapse. “I’d be lying if I told you I knew it would get that bad,” says Braun, now a consultant in business information management at Capgemini. But he looked at industry trends, drawing on data that indicated an increase in inventory (properties languishing in the pipeline) and a slowdown in development. He modeled that information with Cognos Business Intelligence (BI) tools and embedded predictive analytics. Seeing red flags in all that, Braun urged his board to slow its real estate purchasing.

That didn’t go over so well. Braun’s suggestion that real estate prices were not going to rise forever was not popular. “Real estate was showing a great deal of return,” he recalls. “It was a good revenue stream.”

But OmniTRAX did transfer its focus from real estate acquisition to its operating assets. That shift, according to Braun, was a “real lifeline” for the company when real estate blew up, the financial markets tottered, and credit froze.

Many companies and financial executives had access to the same tools as Braun did, and the data was available to everyone. But he had a background in structured finance and securitization, enabling him to see “what was happing in credit.” He added the value of his knowledge and experience to the tools and data.

The Data-Driven Enterprise
In Competing on Analytics: The New Science of Winning, Thomas Davenport and Jeanne Harris define analytics as a subset of BI, “a set of technologies and processes that use data to understand and analyze business performance.” Gartner, which last fall identified next-generation analytics as one its top 10 strategic technologies for 2012, defines it as embedded in technology (as opposed to offline, ad hoc analysis done through Excel), forward rather than backward looking, and drawing from increasingly diverse sources of unstructured information (i.e., Big Data).

Traditional BI uses transactional data captured by an organization’s enterprise resource planning system. New analytics tools draw from Big Data to provide a less transactional, more rounded view of customers and business processes. However, writes Davenport, organizations that win understand that “it’s the human and organizational aspects of analytical competition that are truly differentiating.” In other words, if Braun had read the real estate data differently, or if his board hadn’t listened, the tools he deployed as CFO would not have helped OmniTRAX — not even a little bit.

Davenport writes that “as most analytical projects should involve some sort of financial information or returns, the CFO is at least a partial player in virtually all of them.” David O’Connell, principal analyst at Nucleus Research, claims that the rate of return on analytics investments is $10.66 on every dollar invested. “That means,” he says, “that if you’re gathering data and you haven’t deployed analytics, you’re crazy.”

All That Data
According to a 2011 IDC study, the amount of data being stored by enterprises is doubling every two years. The usual suspects (IBM, SAP, Oracle, SAS, Microsoft) have developed enterprise-grade BI tools to make sense of all that data, much of it unstructured and streaming into the enterprise from the web, connected devices, and social-media feeds in the form of blogs, tweets, “likes,” video, and audio.

A passel of smaller vendors (Lavastorm, Pentaho, Adaptive Planning) offer a variety of database management and analytics tools to work independently of or in concert with the software giants, or (as is the case with software-as-a-service analytics provider Cloud9) with Salesforce to improve the accuracy of deal forecasting. Just as there’s no lack of data, there’s no dearth of tools. But, says Jonathan L.S. Byrnes, MIT senior lecturer and author of Islands of Profit in a Sea of Red Ink, “It’s wrong to start with data. Start by asking what the company should be doing, then reach into the data box, and then apply analytics.”

Advertising is one of AOL’s two main revenue drivers (the other is dial-up service) and therefore what the company should be doing, as Byrnes might say, is use Big Data to inform processes around advertising. AOL deputy controller Cindy Gallagher leverages HP Business Intelligence and Analytics to help her with advertising billing and revenue recognition and also to allow AOL’s sales team to see who clicked on what ad when.

It’s about “understanding your user base to leverage your understanding of behavior,” says AOL vice president of BI Donnie Yancey. With that behavioral information, AOL can optimize ads for specific times and audiences. “My job,” says Gallagher, “is to make sure we’re using the right reporting to get to the sales team and get the right numbers to record revenue and invoice customers.”

On a more mundane but nonetheless important level, last year Gallagher implemented new policies for controlling travel and entertainment (T&E) expenses based on data drawn from the HP tool. Seeing that AOL employees were expensing hotels on weekends (“We wouldn’t have expected that,” says Gallagher), Gallagher implemented rules that made hotels and flights available only on specific days: what she calls “a more financially-friendly T&E policy” that “captured more vendor discounts to drive working capital.”

Recognizing Limits
The promises of analytics, BI, and Big Data are seductive, and success stories abound, but one can’t bring software into a business to optimize every business process. “Optimizing the status quo does not lead to change,” Byrnes points out. “If you optimize something that’s not important, you’re wasting time and money and perhaps deoptimizing what’s really important.”

Defining what’s really important is up to the CEO and CFO, not to tools. Or, as Harvard Business School professor and Balanced Scorecard guru Robert Kaplan noted at CFO’s recent Corporate Performance Management Conference, “Strategy is executed by people, not spreadsheets or BI software.”