In-Depth

HyperRoll and Hyperion Settle Accounts

Old patent infringement lawsuits don’t die, they just sort of fritter away....

• By Stephen Swoyer
• 10/18/2006

Old patent infringement lawsuits don’t die, they just sort of fritter away. Or wind up relegated—once they’re settled—to the back pages of the Wall Street Journal. Unless, of course, they involve a scrappy business intelligence (BI) innovator and an established OLAP giant. Then they’re bona-fide news items.

That was the case with HyperRoll Inc.’s patent infringement lawsuit against OLAP giant Hyperion Solutions Corp. Nearly two years ago, only months after Hyperion announced a new Aggregate Storage Option (ASO) for its Essbase OLAP engine, HyperRoll filed suit, alleging that ASO violated its intellectual property, which it claimed to have shared with Hyperion under NDA. Last week, the two companies settled their differences, forsaking litigation and notching a technology cross-licensing agreement. In addition, Hyperion made an equity investment in HyperRoll for an undisclosed amount.

Much has changed in the past 22 months. Way back then, HyperRoll had carved out a niche for itself as a provider of high-performance OLAP software for Essbase and Oracle’s Express OLAP platform. Hyperion’s ASO effectively spelled the end of HyperRoll’s Essbase niche, however, and Express—which Oracle has long deemphasized in favor of its own integrated OLAP engine—is a terminal market, at best. As a result, HyperRoll officials acknowledge, OLAP is no longer the company’s chief focus.

"It’s not our strategy going forward. It’s not part of our vision. Clearly, we’ll continue to support our OLAP customers. The challenge is that with ADO now in place, Essbase now has the full capability, and Express [usage] is on the decline," comments VP of marketing Eric Rogge.

So what is HyperRoll’s focus? Relational data, of all things.

It’s not nearly so much of a stretch as it might sound. For six years now, in fact, HyperRoll has supported relational data sources, too. It first shipped Relational and OLAP versions ("Suites") of its eponymous performance enhancement technology back in November of 2000. Then, the following year it announced HyperRoll 2.2, a consolidated version of its once-discrete relational and OLAP suites.

Over the next few years, however, HyperRoll tended to play up its OLAP expertise to the detriment of its relational capabilities—possibly as a result of surging sales into Essbase (and to a smaller degree, Express) accounts. (HyperRoll didn’t altogether quash its relational ambitions, however; when it announced HyperRoll 3.0 in October of 2003, for example, it touted that offering’s native access to IBM Corp.’s DB2 relational database.)

It wasn’t until Hyperion’s long-rumored ASO (code-named "Ukraine") loomed on the horizon in early 2004 that HyperRoll again switched gears, touting its ability to supercharge relational performance, too—even though such capabilities had been part-and-parcel of its technology from the get-go. (http://esj.com/business_intelligence/article.aspx?EditorialsID=7110)

These days, Rogge says, his company’s value proposition is the same as it ever was: HyperRoll accelerates query response times and improves transactional performance for a number of different relational database platforms. While the company’s OLAP revenues have stagnated, its relational income—and decision support-related licensing agreements or partnerships (with ProLogic, Ioppolo, and Actuate, among others)—have surged, Rogge maintains.

"Most of our new license revenue has been for our relational [technology]," he confirms. "The big driver for a lot of our customers is significantly improved query response time. Queries in relational databases quite often are burdened by joins, by table scans, by other kinds of necessary operations that a relational database takes just to be a general purpose database. We felt that with our aggregation algorithms—which, by the way, do on-the-fly aggregation, sort of stripping down all of the unnecessary things—that we could respond to decision-support queries much faster than relational databases."

HyperRoll’s aggregation and query acceleration capabilities are especially popular in environments that depend extensively on summarized tables, Rogge says. "In many cases, we’re being used to substitute for summary tables, because they end up having gaps between them for performance. In situations where there’s a lot of ad hoc usage, some organizations are using a summary table strategy to help improve performance," he points out.

As for the merits of its case against Hyperion—which HyperRoll charged with wrecking its business model—Rogge says that’s just water under the all-too-proverbial bridge. It’s not unusual, he says, for patent infringement suits to be settled out of court. And the two companies have firmed up ties—via technology cross-licensing and a minority investment on Hyperion’s part.

"The net of it was that we really didn’t want to get into a situation where we couldn’t openly be partners. That’s the ultimate conclusion, that we decided to end the litigation and initiate the partnership," Rogge maintains. "The details of the partnership are basically that they will be OEM-ing our [relational] technology to be included in projects yet to be announced, and that they will be investing in HyperRoll. They’ve made a minority equity investment, and we’re going to be cooperating on some projects going forward."

In particular, Rogge says, the two vendors have pledged to ratchet up interoperability between HyperRoll’s technology and Hyperion’s System 9 suite.

All things considered, the HyperRoll/Hyperion settlement is a welcome development, says Mike Schiff, a principal with data warehousing consultancy MAS Strategies. Whatever the merits of HyperRoll’s intellectual property claims, the controversy at least compelled HyperRoll to focus on—i.e., more aggressively promote—its languishing relational performance capabilities.

Previously, HyperRoll had concentrated its marketing efforts almost exclusively on OLAP and Essbase—in effect putting all of its eggs in one basket, Schiff says. That isn’t a recipe for business growth, much less survival. "I don’t know how many times I told them that they should stop trying to fixate on Essbase and promote their relational stuff, too," he concludes. "A lot of people have stereotyped them as just working with Hyperion, but the truth is that they do much more than that. They’ve had the ability to work with pure relational databases for many years. They went through this dark period about Hyperion once the suit started, and it wasn’t until then that they really started to focus on the relational side. This [settlement] isn’t a bad move for them, really."