It seems that hardly a week goes by without the introduction of some new way to wrangle insights from data. That breakneck evolution has contributed much of the momentum behind the meteoric rise of analytics in the enterprise, but the constant technological change can distract from ultimate goal of analytics, which is to make more money.

That’s how Bill Schmarzo of EMC candidly put it in his most recent appearance on theCUBE. Profit is ultimately the guiding force behind most commercial data science projects, but that’s not always apparent in the way some organizations approach Hadoop.

Schmarzo, who is chief technology officer for enterprise information management & analytics at EMC, has a first-hand view of the challenges involved in implementing the platform. He said customers often become unnecessarily caught up in the technical details of a project. That’s unavoidable when it comes to emerging technologies, but he noted that the data-crunching framework is hardly that anymore.

“Don’t do a proof-of-concept on Hadoop; Google and Yahoo are already using it. The technology works,” Schmarzo told theCUBE hosts John Furrier and Jeff Kelly. “The challenge organizations have is not with the technology; the challenge organizations have is to ensure buy-in and drive a positive return.”

That’s easier said than done. Properly planning and executing a “proof-of-value” project, as he called it, requires a much more deliberate approach than simply setting up a small-scale Hadoop cluster to test out its capabilities. The journey starts not with the IT department but with the decision-makers who stand to benefit the most from the initiative and have the organizational clout to push it forward.

Of course, explaining the merits of an open-source project for crunching exotic data to business executives and then helping them articulate their goals around that concept can be tricky. So much so, in fact, that Schmarzo said EMC typically spends a full two weeks with customers hammering out the objectives of their Hadoop initiative.

That usually means defining the project in terms of the problem that requires addressing and the questions the end-user needs to have answered in order to solve it. Once that basic outline is established, EMC will create a mock-up of the potential application – such as a mobile app with embedded analytics functions or a new dashboard – to communicate the value to the participant.

That provides a framework to define the specific details of the project. “We try to decompose it into the data events that occur. And there’s almost always some data event where some time limit exists,” Schmarzo elaborated. “The time limit may be measured in minutes or hours, but it’s not batch.”