Cloud and Hadoop? There’s a lot of technology floating around these days and everyone is expected to get on board. However, that often becomes an expensive mistake.
Why does this often become an expensive mistake? Because executives and managers, enamored with the bells and whistles of the latest and greatest tech, plunk down huge sums of money expecting that, once they throw technology on top of their organizations, they will receive solutions that will magically deliver competitiveness, productivity and profitability.
What is needed is not shiny new technologies but rather a transformation in the way organizations are managed. Namely, they need to be open to innovation, tolerant of failure and ready to try new ways of doing things.
The real-time business is a great example of such transformative thinking. Simply buying and setting up a lot of real-time solutions is not going to turn a business into a real-time enterprise. The business case needs to be made and the organization prepared. And, most importantly, it should be implemented in only select processes or areas of the business.
That’s why cloud is such a compelling approach to getting started and building a real-time enterprise. Huge, upfront investments are not required; rather, managers can experiment and pilot their efforts to achieve small-scale success within limited projects and scale from there.
“Keeping initial costs low while building a business case with cloud” was the subject of a recent online panel discussion led by Eric Kavanagh, CEO of The Bloor Group. He was joined by Robin Bloor, Chief Analyst at The Bloor Group and analyst Lyndsay Wise, president and founder of Wise Analytics. Marc Clark, Director of Cloud Strategy and Deployment at Teradata Corporation, also participated in the discussion.
Ultimately, since real-time analytics can be part of so many different parts and processes of a business, cloud provides the flexibility to target such efforts, according to Bloor.
“Analytics isn’t that general,” he said. “When you’re talking about analytics in the cloud, one organization may be particularly keen on creating some kind of predictive analytic capability and injecting what they’re discovering in near real-time into an operational environment.
“Other people may be trying to predict the next activity of their market for up-selling or cross-selling. Others may be just looking at data mining, looking for relationships so they can take advantage of various correlations.
Clark, for one, cautioned against simply adopting real-time analytics for analytics’ sake – a misconception he often sees with companies new to the technology.
“A common refrain is, ‘I’ve been told I/m supposed to use the cloud and [Apache] Hadoop but I don’t know why,'” Clark said. Many midmarket companies don’t have the time, they don’t have the people, they don’t have the resources. He added that the cloud is “not a silver bullet. There are still people that need to be around.
Another mistake companies make in moving to real-time analytics in the cloud is assuming there will be cost savings. It may not save money but it does enable enterprises to move more rapidly with their analytics, according to Clark.
Instead of getting hardware and taking months to get where you’re going,” he said, “you can have a cloud instance up and running and starting to move data within a day. That’s [the] game changer: time to market (TTM).