In-Depth

Hyperion Revamps Essbase, Releases New Analytics

Essbase gets its second coming-out party, and a new name, in nearly as many months

• By Stephen Swoyer
• 09/22/2004

This summer, Hyperion Solutions Corp. announced a new version of its flagship OLAP server—dubbed Essbase 7X—with little fanfare. In fact, in a move that struck some industry watchers as potentially a bid to avoid advance publicity, Hyperion made its announcement the week before the July 4th holiday (http://www.esj.com/business_intelligence/article.asp?id=7155&t=y).

As now seems clear, there was a method to Hyperion’s July 4th madness.

This week, the BI giant finally trumpeted its new Essbase release—along with several other deliverables, including an analytics visualization tool (Hyperion Visual Explorer), new connectivity between Essbase 7X and SAP Business Warehouse (BW), and “Profitability Management,” a bundle that combines analytic software and best practices.

“We actually shipped the software at the end of June, so some [early adopters] could start using it then,” explains Srikant Gokulnatha. Gokulnatha downplays the suggestion that Hyperion wanted a two-for-one PR hit. Nor, he says, is Hyperion hyping the not-so-new Essbase to call attention to its other new offerings.

At least one thing did change between late June and mid-September: the new Essbase’s branding. Three months ago, Hyperion called it Essbase 7.1; it’s since been rechristened Essbase 7X.

Regardless of its name, Essbase 7X ships with Hyperion’s much-anticipated Aggregate Storage Option, a technology that can, in some cases, dramatically accelerate the performance of the Essbase OLAP engine. In this respect, Essbase 7X’s new Aggregate Storage Option addresses an issue—storage of sparse dimensions in an OLAP data cube—that’s been a bread-and-butter niche for Hyperoll Inc., a vendor that markets an add-on product for both Essbase and Oracle Corp.’s Express OLAP platforms.

“What we’ve done is we had an existing storage layer that was very optimized for storing dense data sets for financial applications, and we called that the block storage layer. What we’ve done with Essbase 7X is add another storage layer that fits in parallel with the dense storage layer and that’s highly optimized for sparse data sets, so we also provide the opportunity to use both of those together,” Gokulnatha explains.

Gokulnatha and other Hyperion officials, of course, don’t even so much as allude to HyperRoll. Instead, they argue that the Aggregate Storage Option and other new performance enhancements are just part of Hyperion’s strategy to make Essbase all things to all OLAP consumers.

“You have technologies that are designed to focus on financial analytics, and … on the other side, we have applications that focus on delivering analytics for sales, marketing, manufacturing—operational analytics,” he says, arguing that what’s needed is a hybrid approach, enterprise analytics, that can address the distinct needs of both types of analytic consumers equally well.

Not surprisingly, Gokulnatha says, Essbase 7X is designed for just such a purpose. The new Essbase release also appeals to another trend—infrastructure simplification—that’s becoming increasingly popular: “Companies are standardizing, or would like to standardize, on as few products as possible. Very simply for TCO reasons, they want to take one technology that addresses all of their needs, and they want to learn one development platform, and obviously that simplifies their relationships as well.”

At the same time, Gokulnatha stresses, Essbase is already widely used to support operational analytics. “Literally hundreds of our existing customers use Essbase for operational analytic applications, so there was nothing that prevented this. The key was to do this efficiently when customers tried to scale data, especially with a large number of dimensions,” he comments. “They can now [scale] without any design compromises, without any worry, use that same product for literally any application in their enterprise.”

Mike Schiff, a senior analyst with consultancy Current Analysis Inc., has described HyperRoll’s approach to sparse data sets as a “black box thing”—largely, he says, because that company has declined to disclose exactly how its technology works. Nevertheless, Schiff acknowledges, HyperRoll isn’t selling snake oil. Nor, he speculates, is Hyperion.

“For example, not every store sells every product in every color everyday, but in an OLAP cube, you have to think about that, which requires extra time and extra processing power. HyperRoll seems to address [such scenarios] and makes [the OLAP engine] a lot more scalable, and my guess is that’s what [the Aggregate Storage Option] does, too. The idea is that you should be able to both increase performance and handle larger cube sizes.”

Hyperion says that several customers, including retail giant Belk, are already using the revamped Essbase 7x. “The example I usually show is an application where the cube creation time went from around three hours down to just under two minutes,” Gokulnatha confirms.

Elsewhere, he says, the new Hyperion Visual Explorer is the product of several years of R&D. It’s designed to let customers more easily (i.e., with point-and-click, drag-and-drop simplicity) build analytics.

“It provides a way for end users to use a drag-and-drop approach onto a visual canvas to very quickly create analytics in a visual fashion,” Gokulnatha explains. “The tool is smart enough to recognize the asset the end user is operating on, and the kind of analysis they’re trying to do”—for example, trending analysis.

Because Essbase is currently used extensively in SAP ERP environments, Hyperion officials say, many customers have specifically requested that Hyperion improve integration between Essbase and SAP BW. “There are two drivers here,” says Gokulnatha. “The analytics they can do in the Essbase world are richer and more complex than what they can do in SAP BW. The performance within SAP BW is not the fastest, and by moving a subset of the data into Essbase, you can improve your analytic performance quite dramatically.”

Finally, Gokulnatha says the new Profitability Management bundle taps Hyperion’s years of expertise as a purveyor of financial analytic applications. The new offering taps a combination of reports, dashboards, and analytics, along with allocations and operational models, to help executives make decisions about which customers are profitable or unprofitable, margin contributions of different products; specific customer purchasing behaviors; the cost of acquiring new customers, and etc. “What we’re doing here is bringing those tools together. We decided to package together the software that we have with the best practices and the [intellectual property] we have built up over time."