In-Depth

ETL Competition Fierce for Mid-Market Products

Informatica, SAS, and Ascential control the high-end ETL market, but the mid-market is still largely up for grabs as the low-end market shrinks.

• By Stephen Swoyer
• 08/18/2004

According to consultancy Forrester Research, three vendors—Informatica Corp., SAS Institute Inc. and Ascential Software Corp.—collectively account for more than half of all ETL-related revenues.

But while these three vendors effectively dominate the high-market segment of the overall ETL space, Russom says, they face fierce competition in the mid-market, where ETL-only revenue can range between $20 million and $40 million per vendor. Mid-market ETL vendors include DataMirror, IBM, Microsoft, Business Objects, Ab Initio Software, Computer Associates, Oracle, and Cognos.

Similarly, high-market leaders Informatica, SAS, and Ascential haven’t, as of yet, made strong moves into the low-end ETL space—and with good reason, if Russom’s market research is accurate. That’s because the low-end ETL market segment—which once accounted for 19 percent of all revenues—is shrinking, Russom finds, with as many as 20 established vendors vying for customers.

“Most products in this segment were hit hard by the drop in ETL spending of 2002 and they did not rebound in 2003 as gracefully as products from other segments,” he writes.

While low-end ETL revenues continue to stagnate, mid-market vendors are posting average growth, and high-market vendors continue to grow their revenues at an impressive clip. “In the ETL market, the big are getting bigger, and the small are mostly staying small, due to annual growth of 5 percent or less,” he writes. “Between these extremes, most products of the middle segment are growing revenue at the market-wide average of 10 percent.”

Nevertheless, says Russom, there’s an opportunity for low-end and mid-market vendors to capitalize on an as-yet-untapped market segment. That’s because there’s a substantive gap between the revenues of mid-market vendors – which typically max out at $40 million per vendor – and those at the high-end. “[T]here is no vendor with ETL-only revenue in the $40M-$80M range (2001-2002) or $40M-$120M range (2003),” he writes. “The three leading vendors of the high-market segment have cornered the market, and are 300 percent larger than their largest competitors of the middle segment. Since high-market growth is better than that of other segments, the gap will continue to increase.”

One strategy for low- and mid-market vendors, says Russom, is to target this gap. “Many user organizations are unwilling to pay for an expensive high-market tool, but are willing to consider an alternative,” he writes. “Vendors should consider filling this void with a tool that meets most of the enterprise-scale requirements of the high segment, but at a price point set between high and middle segments.” Several mid-market vendors—including Ab Initio, Business Objects, Microsoft, and Oracle—are well positioned to exploit this gap, he says, stressing that these vendors must first address feature, functionality, or scalability problems to make any significant headway.

What about outside vendors that are angling to get into the ETL space? Business Objects, for example, became an established ETL player almost overnight when it acquired the former Acta Technologies several years ago. Ditto for Pervasive, which acquired the former Data Junction last year.

For outside vendors who are hoping to buy their way to ETL glory, Russom has some sobering news. “The answer is that very few pureplay ETL vendors are left, because the last good ones—Acta, Data Junction, and Sagent—were acquired when market caps fell early this decade,” he writes. “On one side of the coin, vendors hoping to enter the ETL market any time soon should build a new tool. On the flip side, with demand for acquirable ETL startups rising, software entrepreneurs and venture capitalists should look seriously at the ETL market.”

That’s not to write ETL pure-plays off altogether, however. Russom believes that several emerging technologies or requirements, such as radio frequency identification (RFID) and regulatory reporting, will encourage a raft of new pureplay ETL vendors. “Small vendors have traditionally been a source of innovation, bringing to market once leading-edge capabilities for realtime, packaged connectors, extreme scalability, customer data functions, and so on,” he writes. “Eventually the pure-play pool will repopulate with new startups addressing emerging requirements. … And then we’ll again go through the age-old cycle of these small vendors either growing to midsize or being acquired.”

On the whole, Russom forecasts a rosy future for ETL market leaders Informatica, SAS, and Ascential—as well as for the ETL market, as a whole, which he expects will reach the $1 billion milestone in 2005.

Not everyone agrees, however. Take Mark Smith, CEO and senior vice-president of research with consultancy Ventana Research. Where Russom forecasts at least 10 percent annual ETL growth through 2006, Smith foresees an ETL-buying chill in 2004—at least for data integration pureplays Informatica and Ascential. “While I think the plumbing is important and [the ETL market experienced] good growth in 2003, this year has been a spoiler for Informatica and Ascential,” Smith says in an e-mail, citing the poor financial results of both companies. “Many organizations have put a hold on new ‘plumbing’ until they can assess where and why the pipes need fixing. [ETL is] still a growth market opportunity, but for now the buying power on this stuff has cooled off.”

He also takes issue with Russom’s suggestion that increased usage of radio frequency identification (RFID) will contribute to an upsurge in demand for ETL. “On RFID, if people use ETL for this, we are all in trouble—event and message-based integration will need to handle most of this approach."