New customers drive Teradata to new heights
Big data opportunity a big unknown
When announcing its financial performance for its most recent quarter, data-warehousing pioneer Teradata proved that it's continuing to do well despite heavy pressure from IBM, Oracle, EMC, and HP – and also, like its rivals, that its still trying to figure out just how large of an opportunity business analytics on unstructured data can be.
Teradata dabbles a bit in the big-data arena with its partnership with Cloudera, a peddler of a commercially supported and extended version of the Apache Hadoop MapReduce and distributed file system that mimics an earlier style of computing done by search engine giant Google.
However, for the moment, Teradata's data warehousing systems – be they large-scale Enterprise Data Warehouses or smaller appliances with various names – chew on structured data stored in the company's eponymous distributed, clustered database-management system, and are generating a widening stream of revenue.
In the second quarter ended in June, Teradata's revenues were up 24 per cent to $581m, accelerating from the first quarter. Net income was up 39 per cent to $103m, and Teradata is bringing 18 per cent of revenues to the bottom line, which is not bad for a hardware company.
Product sales rose by 21 per cent to $269m, and in a conference call with Wall Street analysts on Thursday, Teradata president and CEO Mike Koehler said that during the first half of the year, the company booked the highest number of new customers that it has seen in the past decade, with companies in the financial services and retailing businesses driving a lot of the revenue across both new and existing customers. Teradata doesn't break out product sales by line, but Koehler said that Teradata sold a "significant" number of its hybrid disk-flash appliances in the quarter.
Maintenance services were up by 17 per cent to $135m, while consulting services rose a very decent 34 per cent to $177m. New account sales are, in part, driving those consulting sales.
"In a new account, there is a higher level of consulting services in the mix compared to the hardware sale, versus an upgrade in an existing account," explained Koehler. He added that Teradata was building up its consulting practice and was seeing more companies coming to it to sort out their data-analytics needs because of its expertise, broader product line, and experience. Overall, services revenues in the quarter were up 26 per cent to $312m.
Aprimo, a Web-based marketing services company that Teradata acquired in December 2010 for $550m and closed in April, contributed a little to the kitty this time around. It's also on track to contribute more, thanks to cross-selling, in the future. Koehler said that bookings for Aprimo were up 30 per cent year-on-year in the wake of the acquisition, and for the year, Aprimo is expected to add about 3 per cent to Teradata's top line.
Aster Data, which created an SQL-MapReduce hybrid and which Teradata bought in March of this year for $263m, is not expected to add much to the top line this year, but Teradata has high hopes for it.
The one thing everyone is trying to figure out right now is how big the "big data" addressable market is – and guess what? Teradata doesn't know, either. (Maybe they should have shelled out $1bn for SPSS for its predictive analytics, instead of IBM, back in July 2009.) "At this juncture, I don't think the market opportunity is clear, but it is large – extremely large," Koehler said in the call.
What about competitive pressures, with Oracle and IBM being very aggressive in data warehousing? What did Teradata see in Q2? "Other than some advertisements, not a lot," Koehler said with a laugh.
"I have never seen our competitive position so strong. Our competitive win rate really started to kick in a year and a half ago. Since Netezza [now part of IBM] entered the market three years ago, our win rate has gone up. We compete more with Oracle, but we really can't find much Exadata in our accounts. The only thing we can figure is they are mainly used with OLTP."
This is what Teradata has been saying for several quarters, and Oracle doesn't provide specifics about its Exadata machines – which can in fact be used to support data warehousing or OLTP workloads, unlike Teradata and Netezza iron, which only does data warehousing.
Teradata had planned to add 30 new sales regions this year, focusing on specific industry verticals, mid-market customers, and growing and emerging markets in Europe and Asia, and as of the end of June, all of these were up and running. Koehler said that Teradata would therefore add another 30 regions to help push sales. The added 30 regions have already contributed more than $50m in revenues during the first half of 2011. These regions take three or four years to hit their full yield level, Teradata says.
Because of the need for more data analytics and interest in big-data munching, Teradata is raising its revenue guidance for 2001 by 4 per cent, and now expects sales to rise by between 18 and 20 per cent. The company also raised earnings-per-share guidance by a few pennies, to between $2.20 and $2.28 per share. The expansion in those additional 30 territories will ding profits for a bit, but will eventually yield a stronger bottom line over the long haul, says Teradata.
As El Reg goes to press, the Dow Jones Industrial Average is down 4.3 per cent, – 512.5 points – but Teradata is only off a fraction of a per cent because of its good numbers and raised guidance. It was one of four companies that actually had a stock that was up – initially by 6 per cent early in the day – in the S&P 500 during a trading day that deteriorated as the day wore on because of nervousness about the global economy.
Even with that tiny haircut, Teradata is way too pricey at a market capitalization of over $9bn for anyone to consider buying. Spinning out Teradata might just go down as one of the dumbest things NCR has ever done. ®
You may have a point
When we're running several terabytes at home, and the data warehouse at work is pedalling a couple of zettabytes...
"Spinning out Teradata might just go down as one of the dumbest things NCR has ever done"
Whats in a name?
I can't help but think Teradata as a name is is going to seem a bit dated in 10 years or so.