Netezza IPO: Stacked Up Against the Big Guns

Netezza (NTZA) (originally called Intelligent Data Engines), the data warehouse [DW] appliance supplier, filed its IPO on March 22. Because I analyzed the buzz around the “software” appliance concept recently, let’s apply the same methodology to the DW appliance market opportunity as Netezza describes it in its offering documents. The analytical methodology goes like this:

• Is the IT market revenue opportunity incremental or a zero-sum scenario?
• Even if a zero-sum scenario, which supplier or suppliers can gain in the scenario and which suffer?

With this process, you quickly realize that Netezza is up against the big guns, not the DATAllegro’s, Dataupia’s, and Greenplum’s that conventional wisdom compares it against.

In terms of whether DW is an incremental IT market revenue opportunity, I think the “need” to store and then analyze petabytes upon exabytes of transactional data in near real time to support homeland security, various compliance laws (pharmaceutical, financial services, accounting, etc.), scientific inquiry, and web-based sales and marketing is an incremental IT market revenue opportunity. (“Need” is in quotes because you can have some good philosophical and political discussions about this sort of “big brother is watching us” use of technology but at this point, whether you like it or not, it’s a good bet that “he” is.)

However, there are both multiple ways to accomplish massive real-time data storage and analysis (of which DW is only one), and multiple ways to bring DW or its alternatives to market. For this reason, it is not good IT market investment research to think that there is such a thing as a “real-time DW appliance” market. That’s slicing the market opportunity too finely. There is a good explanation of the technology underlying the various approaches in this DM Review Magazine article (truth in advertising: the author of the linked-to article is or was employed by a company with a competitive concept to Netezza’s).

Netezza is competing in the same market as suppliers selling DW software functionality via a license or as a subscription service. Netezza is also in the same market as suppliers offering DW virtual software appliances (whereas Netezza’s appliance is hardwired). In addition, the other techniques of massive real-time data storage and analysis include analyzing the data right in the transactional database or possibly subsetting the historical data into data marts. Netezza does not compete for those opportunities at all.

I believe Netezza competes for DW opportunities only (1) because that is how it describes itself and (2) because the Netezza IPO quotes an IDC estimate that the DW management software market alone will “increase from $4.0 billion in 2005 to $6.3 billion in 2010, a compound annual growth rate of 9.6%.” That is healthy growth but not spectacular remembering that the IDC numbers are most likely calculating both new opportunities and the revenue from maintenance fees on the thousands of DW-software installs over the last 15 years.

IDC is also most likely measuring Windows-operating-system-based DW opportunities along with opportunities where UNIX, Linux or legacy operating software is deployed as the infrastructure. Netezza does not compete for the Windows-based opportunities [unless it can convince an enterprise to change an explicit or implicit Microsoft (MSFT) strategy], further limiting its overall revenue base. I can’t be sure exactly what the IDC forecast measures because the underlying IDC document is not referenced in the Netezza IPO.

As a further proof point about the steady but slow nature of growth in the DW software market, NCR (NCR) Teradata DW revenue increased 6% in 2006 from 2005 (foreign currency fluctuations had less than 1% of negative impact on the year-over-year revenue comparison), slowing down from 9% growth in 2005.

In its IPO, Netezza adds to the IDC factoid with its opinion that the software required to manage a DW represents only a portion of the DW’s total cost. Netezza says warehouses also require hardware such as servers and storage as well as services to install, integrate and manage the DW. What Netezza is saying is that its “DW appliance” packages two of those three ingredients and that it—Netezza—also offers related installation services. That’s true, but typically the market growth rates for those ingredients track the software market’s curve.

So, answering my first analytical question, massive real-time data storage and analysis is an incremental revenue opportunity (because more and more data needs to be warehoused). But the market for DW appliances is not an incremental opportunity; it is ‘just’ part of a bigger steadily but not spectacularly growing DW market, which—in turn—is only part of the incremental opportunity. So Netezza’s real-time, appliance-packaged DW is competing in a zero-sum-scenario.

Who wins and loses in such a zero-sum scenario? Typically big, established players win. Often (but not always) a startup such as Netezza or a restart breaks out of a pack of other startups and inserts itself into the mix to becomes a big, established player. Apple (AAPL) and Microsoft in the 70s, Oracle (ORCL) and SAP (SAP) in the 80s, EMC (EMC) and Sun (SUNW) in the 90s, and Google (GOOG) so far this decade are all examples among today’s IT Top 12. Wang in the 80s, Peoplesoft in the 90s, and Siebel this decade showed similar growth patterns for as many as 10 years but of course did not survive. [If you think Netezza might be a Wang, Peoplesoft, or a Siebel, look at it as an investment for its take-over potential (hard to do when it is still pre IPO but not impossible). As with many startups, Netezza’s business plan might be to be taken over by one of the above. EMC, whose corporate headquarters is a few miles from Netezza’s, would be a likely acquirer in such a scenario.]

Netezza says it brings two new wrinkles to the market:

• almost real-time access, whereas earlier generation queries against DWs took hours or days
• “appliance” packaging, whereas earlier generation DWs required the same level of systems, network and database administration as any transactional application

But be careful of these claims; perhaps Netezza can enable these types of improvements but so can the big established players very quickly. IBM (IBM), in a sense, already has.

At its highest level of functionality, the Netezza Performance Server [NPS] is basically a development tool. The IPO says Netezza provides a business intelligence [BI] solution but the product does not provide the BI application itself; it simply enables it. The risks section of the Netezza IPO is clear about this when it says:

Our (Netezza’s) success depends on the continued recognition of the need for business intelligence in the marketplace and on the adoption by our customers of data warehouse appliances, often as replacements for existing systems, to enable business intelligence.

The startup/big-guy difference and the solution/enabler distinction is not necessarily a bad thing for Netezza but it changes its comparatives from the DATAllegro’s, Dataupia’s and Greenplum’s that Netezza is usually associated with to IBM, Oracle (ORCL), and maybe Sun (SUNW) to a lesser extent. NCR is clearly a comparative and it will be more of a direct competitor after NCR completes the Teradata spinout. (EMC (EMC), HP (HPQ), and Sybase (SY) are also called out in the Netezza IPO but I do not see these companies putting an emphasis on DW at this time).

Can Netezza displace any of these big guys in the long run in the DW market? It needs good references and a probably a flawless channel strategy execution (because of the deep, wide direct sales forces of IBM, NCR and Oracle). It (nor can any startup in any market) rest on product functionality alone.

In terms of customers, in its IPO Netezza lists some good customers signed up already. I am not sure I would have listed the VA and Army Corps of Engineers as customers; hopefully the Netezza systems are going into those agencies to solve the problems we are hearing about, and didn’t cause them. The U.S. federal government accounted for about 5% of Netezza’s revenue in FY 2007. One customer that was highlighted early on in Netezza’s history was TJX (TJX), based in the same city as Netezza. Again, hopefully, the Netezza system was not the enabler of the well-publicized data identity thefts at TJX and instead will be a fix for those problems. (Netezza would not comment on my question concerning either possibility; TJX did not return my call.) The Netezza appliance is a big ticket item so the market potential is limited to the very largest organizations. In 2004 (its fiscal year ended 1/31/2005), 15 customers spent $36 million. Two years later, in 2006, 87 customers spent $80 million. Per-transaction value is coming down as would be expected but we are not talking minor commitment here.

Most importantly, I don’t see a break-out channel play, a key to big success, in general in the past when a company came along to beat the big guys (best examples being Microsoft, Oracle and Sun). Netezza says its future revenue growth will depend in part on the continued development of its indirect sales channel to complement its direct sales force, which accounted for more than 80% of Netezza revenue in 2006. Netezza foresees the creation of a channel called analytical service provider but these are mailing houses by another name.

If analytical service provision is a separate market, Netezza might want to think of entering it itself, analogous to the way Google—early in its corporate existence—decided to change from licensing its search engine in the traditional fashion to selling advertising with free searches.

Although a “propriety product” in terms of both hardware and software, there is an open source software [OSS] sidelight to the Netezza roll out. Netezza makes extensive use of OSS. It says it takes steps to ensure that its proprietary software is not combined with, or does not incorporate, OSS in ways that would require its proprietary software to be subject to an open source license. On the hardware side, Netezza depends on single source suppliers in multiple ways and presumably any change in these relationships would at a minimum cause Netezza to hiccup:

• Netezza has contracted with Persistent Systems Pvt. Ltd., located in Pune, India, to employ a dedicated team of over 50 engineers focused on quality assurance and product integration engineering.
• Netezza relies on a single contract manufacturer, Sanmina-SCI Corporation to assemble its products, manage its supply chain and participate in negotiations regarding component costs
• HP provides some of Netezza’s service capabilities (and of course is also a competitor)