Hewlett-Packard Co.’s data warehousing and business intelligence (BI) strategy took yet another twist this week with HP’s announcement that it is buying analytical database developer Vertica Systems Inc., a deal that creates new uncertainties for vendors and users alike in the data warehouse database market.
The disclosure of the planned acquisition came just three weeks after HP confirmed longstanding rumors that it was ending sales of its own
Buying Vertica raises obvious questions about HP’s commitment to the partnering approach, which also includes an earlier deal with SAP AG. Shawn Rogers, an analyst at Enterprise Management Associates Inc., said at the TDWI World Conference here that HP “seemed to be on a hardware and services trajectory” after scuttling Neoview. Now, he added, “it will be interesting to see how all of this comes together strategically.”
In an interview at the conference, Giuliano Di Vitantonio, director of global marketing and alliances for HP’s Enterprise Information Solutions group, insisted that Vertica’s columnar database “is absolutely not a replacement for Neoview.” The latter was an enterprise data warehouse database, while Di Vitantonio described Vertica as a real-time analytics platform. Partnering “is still our strategy for enterprise data warehousing,” he said. “We don’t see that as the natural market for Vertica.”
A new analytical database start for HP – or more of the same?
Asked why HP thinks it can do better with Vertica than it did with Neoview, Di Vitantonio declined to comment in detail, saying only that HP officials clearly “are extremely optimistic about the prospects, or otherwise we wouldn’t be doing this.”
Di Vitantonio wouldn’t even discuss how HP will use the Vertica software to differentiate itself from other analytical database vendors. “We’re not going to comment on that now, because frankly there’s nothing to comment on,” he said, noting that the acquisition has yet to be completed.
Privately held Vertica claims to have more than 300 customers, and Rogers and other analysts view its database as a solid technology – certainly compared to Neoview, which “was an OLTP DBMS retread that was unproven and had no installed base,” said Wayne Eckerson, research director for TechTarget Inc.’s business applications and architecture media group. Buying Vertica shows that HP “really is serious about BI and analytics” despite Neoview’s failure to make a dent in the market, Eckerson added.
But HP’s unsuccessful track record with Neoview and various other software products leaves doubts about its ability to grow Vertica’s business.
“I think Vertica is a good technology to pick up,” said David Stodder, research director at The Data Warehousing Institute (TDWI), which organized this week’s conference. “It’s just going to be whether HP can make something of it, or if it’s going to fall victim to the same problems HP has had [in the past].”
In addition, Vertica’s software is still maturing as an analytical database, said Ventana Research analyst David Menninger, who was vice president of marketing and product management at Vertica until last June. If HP officials expect the software to generate substantial amounts of revenue “without significant investment [in product development], I think they’re being shortsighted,” Menninger said.
Consolidation comes in threes in the data warehouse database market
The HP-Vertica deal also raises new questions for the remaining independent vendors of analytical databases as well as for users who are looking at the available technologies. Two other rivals already were bought by major IT vendors last year: EMC’s acquisition of Greenplum started the consolidation wave last July and was followed two months later by IBM’s purchase of appliance vendor Netezza.
Executives from Vertica’s competitors put on brave faces at the TDWI conference. John Thompson, CEO of U.S. operations at massively parallel processing (MPP) database vendor Kognitio Ltd., acknowledged that his company “had conversations” with HP prior to this week’s announcement. “We all knew HP was going to buy someone with the demise of Neoview,” Thompson said.
But he added that he expects the acquisition of Vertica to be “a wash” with prospective customers. “It continues the trend of consolidation in the industry, but it doesn’t materially change things,” he said.
Like Thompson, officials from ParAccel Inc. and Aster Data Systems Inc. contended that Vertica’s software is geared to smaller installations than their databases can support. “Sure, Vertica and us are both columnar databases, but we’ve gone at it more from the enterprise level,” said Barry Zane, ParAccel’s chief technology officer.
Standing out versus standing alone on analytical databases
Steve Wooledge, senior director of marketing at Aster, made a similar point about his company’s MPP database, which is targeted primarily at so-called Big Data analytics uses and got added support for column-based storage last September. Wooledge also said that as one of the independent vendors left in the market, “Maybe we’ll stand out more now.”
But Menninger thinks the ongoing consolidation could eventually make it harder for small vendors that don’t find buyers to continue to thrive. “In the long run, it’s a game of musical chairs,” he said. “At some point, the music is going to stop and the independent vendors who remain standing are going to have a harder time maintaining high growth rates.”
At the least, it’s important for those vendors to build up their installed bases in order to remain competitive, according to Rogers. He said the HP-Vertica deal “puts a line in the sand between companies that have large customer bases and those that don’t.”