In introducing its BigQuery service, Google emphasized the small- and medium-sized businesses it says will benefit from a cloud-based analytics service. Companies already using the service during a trial period, like Claritics, which tracks the behavior of online gamers, and Crystalloids, an analytics firm that examines booking patterns for a European resort client.
Both companies wanted to get real-time insights from data queries. But they are not the kinds of companies that need the MapReduce distributed cluster computing technology that Google designed. Instead, they are looking for cloud-based computing power to analyze large quantities of data quickly. And they do not need to modify the data they are analyzing.
BigQuery is poised to compete with established cloud-based offerings such as Amazon Web Services.
Google launched its BigQuery cloud-based data storage and analysis service on May 1.The service offers up to 100 gigabytes of free storage. After that, users pay 12 cents per gigabyte per month up to two terabytes (custom services are available for bigger data sets). The queries run on a SQL interface, and cost $0.035 cents per gigabyte searched.
Ju-kay Kwek, Google’s product manager for BigQuery, said to date analyzing big data sets has been in the domain of large enterprises that can afford the tools and systems, and that’s what Google is out to change.
“There is just a huge amount of capital investment that any business would have to make in the past in order to gain the insights from big data,” he said. “They don’t have to build a data center anymore. (We’re offering) huge scalable big data analysis more readily available to businesses through the cloud.
“We definitely see a lot of customers coming out of the woodwork who are coming to us with this need to capture increasing amounts of data, but it’s not enough to just capture it,” Kwek said. “They want to do interactive analysis, they want to explore the data without knowing how ask the right questions beforehand.”
John Brantley Hooks IV, a big data researcher at GLG Research, said BigQuery “is not a revolutionary step forwards for cloud storage or big data,” but it does place Google at the intersection of those two marketplaces.
“It has also ‘socialized’ big data for smaller companies that need to avoid up front (capital) costs, and those with variable and short term data analytics requirements,” said Hooks, who also consults about cloud services for large financial services companies. “Due to performance requirements I would not see it making inroads into the high-frequency trading market or within many of the financial services markets.”
While the offering should cut out infrastructure costs, companies will need someone with application development experience to develop a front end tool that will make querying data stored by Google easier.
“BigQuery is a software service,” Kwek said. “It is a developer-facing product, so you really want to have somebody who has a developer experience to be able to use the APIs or to be able to use the SQL interface.”
Kwek said the team that developed BigQuery has “worked really hard” to make the product integrate with other offerings, such as Google spreadsheets, to increase its ease of use.
Dave Schubmehl, a research manager at IDC, said the other big question will be companies’ willingness to store large amounts of business data at Google.
“I have my doubts about whether they’re actually going to get people to take their data and actually upload it or make it available to Google BigQuery to be analyzed,” Schubmehl said.
Kwek said the Google is committed to make security a top priority, building it into multiple levels from user access to transmission “over the wire” to security in the database infrastructure. “There is a very conscious effort to build security into the entire stack from the very beginning,” Kwek said.