Big data and the need for speed

The database can be a bottleneck in organisations' efforts to gain actionable insights from their data, says EXASOL CMO

The message that data is an untapped oilfield has been hammered home so relentlessly that there can be few organisations of any size that haven't at least thought about how they might wring some value out of all the information they are sitting on.

Companies understand the potential now, says Sean Jackson, CMO at analytics database management company EXASOL, but that doesn't mean they know how to go about converting data to value.

Too often, he says, companies will just go out and buy a Hadoop cluster without really knowing how they're going to use it. About one-third of EXASOL's new customers are Hadoop users looking for help to get to the next stage.

"There has been a lot of hype about Hadoop, so much so that many see Hadoop as the panacea to their big data problems. It isn't, it's just one part of the solution," says Jackson, who previously worked at big data companies Actian and Kognitio. "Companies want to make use of the data in Hadoop, analysing it and turning it into value as actionable insight."

Others see no need for Hadoop. They just want their business intelligence systems to run faster and to be easier to use. Many have invested in visualisation systems in order to democratise analytics, but business users, who arguably have the most to gain from the insights made possible, are intolerant of having to wait around for the results of their queries.

"What's the point of a glamorous chassis and body if the engine does not perform?" asks Jackson. "Some companies have invested in great front-end BI visualisation tools but need them to run faster, where queries are reduced from minutes to seconds, sometimes sub-seconds."

The need for speed in a data-centric world is perhaps EXASOL's trump card. According to the independent benchmarking organisation TPC, the company's EXASolution in-memory analytical database recently beat all other contenders in terms of performance in tests designed to be representative of real-world workloads.

Not everyone needs a super-fast analytical database of course, but Jackson predicts that performance will become more important as the Internet of Things comes of age.

"As we become ever more interconnected and where dumb devices become smart devices emitting data and information even more, so the need for fast data processing and analytic technology will grow," he says. "The data tap has been turned on; it will never be turned off."

For now, though, most of EXASOL's customers are those for whom "data is the lifeblood of the business".

"It's start-ups who see the intrinsic value in data to build new products and services and who just 'get it'; it's gaming, retail, e-commerce, digital and telco - but less the leading brands where they have invested in mainstream technology over the years, more the smaller companies who want to be more agile, and who want to innovate."

The ability to spin up instances rapidly and cheaply makes cloud platforms popular with small innovative companies, but with data of such core importance, some are wary of cloud computing, in spite of its obvious convenience as an elastic processing platform.

"Every customer has different needs and requirements," Jackson says. "Some want to run in the cloud, but the cloud has to meet their needs, especially when it comes to data security. That's why we work with local cloud partners specific to the country the customer operates in."

He continues: "The cloud offers a great way to get started with analytics without huge up-front investments, but I would question how many businesses operate long-term in the cloud. Longer term, many systems that are in production are brought in-house."