How is Hadoop like teenage sex? It’s an old riddle whose answer is changing quickly. If you don’t already know what it is, read on. And if you do, read on anyway (and check out the Infographic) because we have some marvelous visual statistics to share.
How Is Hadoop Being Used? Survey Says …
During Strata and Hadoop World last month, Hortonworks, a Hadoop platform provider, conducted a survey. It asked 500 conference attendees about how they were adopting Hadoop. A year ago the answers would have been“we’re here to see what it is” or “we’re doing a pilot.” This year, half of the respondents said they were adopting it to identify new business opportunities, the other half said that they were exploring how it would augment their current data architectures. They all had Hadoop-related plans.
“There’s no lack of interest in Hadoop today,” said Jim Walker, Hortonworks’ director of product marketing.
What Kind of Data Belongs in Hadoop?
There were six types of datasets respondents said they were looking at: clickstream data, server log data, social media data, machine data, geolocation data and unstructured data trapped in PDF files, videos and other types content files.
“They are doing really cool things with Hadoop,” said Walker.
While we’re all familiar with how marketers use data stored in Hadoop to bombard us with offers as we surf the web, we probably don’t know that mega law firms are looking at big data to find common denominators in PDF files and to then leverage them to speed the creation of new documents. Or that machine data is being gathered from things like HVAC systems, elevators, thermometers and thermostats and that predictive analytics are then being applied to forecast and prevent breakdowns.
Consumer goods companies are also leveraging big data to not only analyze sentiment about themselves, but about their competitors as well. And they’re even gathering data from places like LinkedIn to learn what kinds of people companies are bringing on board (or what kinds of people are quitting) in order to predict industry trends and to identify sales opportunities.
Not an All or Nothing Game
Companies are also using Hadoop to make their existing applications smarter; in other words, big data is augmenting existing enterprise solutions based on technologies like Teradata, Microsoft and SAP.
Hortonworks has a strong belief that most enterprises aren’t going to ditch their existing systems and go all-in on Hadoop. “You have to be pragmatic,” says Walker. He says that in the short term (five plus years) many enterprises will use Hadoop to inform existing systems (as opposed to replace them) and continue to leverage the skill sets of existing employees.
The Answer to the Riddle
How is Hadoop like teenage sex? The answer used to be, “Everyone is talking about it, but no one’s doing it.”
By this time next year it will be, “Everyone’s doing it and having a great deal of fun.”