Last week Teradata announced its intention to acquire Aster Data for $263 million. This equates to 89% ownership of Aster Data; Teradata acquired an earlier 11% ownership interest back in September last year. The deal, which is expected to close some time in the second quarter, is designed to give Teradata a firmer foothold in the ‘big data’ analytics market.
Riding the ‘big data’ wave
‘Big data’ is a term that aligns itself with exploding data volumes originating from complex and diverse data sources such as blogs, sensors, scientific data, social media platforms and electronic networks . It’s not just the sheer volume of data that poses a challenge for organisations, but being able to leverage all of it to drive out meaningful insights for a business. Although Teradata is often regarded as the final word in high end Enterprise Data Warehousing (EDW), its name had not become synonymous with the current big data movement. Aster Data, on the other hand, was one of an early crop of vendors which took advantage of the big data trend by developing an analytic database that allowed the processing and analysis of massive amounts of diverse and predominately internet-based data. By acquiring Aster Data, Teradata will claim a stronger foothold in the big data camp—where the companies’ combined capabilities will be able to be leveraged to process and unlock intelligence in both traditional and newer expansive and more complex data sources.
There are many parts of Aster Data’s offering that will complement Teradata’s core data warehousing business. As we recently outlined in our report Aster Data: where big data meets analytics, Aster Data‘s ncluster database provides a Massive Parallel Processing (MPP) hybrid row and column analytical database that takes advantage of the MapReduce framework for highly parallel and scalable data processing. This differs somewhat from other big data approaches as it runs MapReduce in-database and combines it with SQL so developers can take advantage of MapReduce’s ability for distributing the processing of large data sets within standard SQL.
The benefit of such an approach is that it allows organisations to build and deploy sophisticated analytics that would otherwise be very difficult to express and query in standard SQL. Aster Data’s implementation of MapReduce and its pre-packaged library of SQL-MapReduce functions make it particularly suitable for analytic applications that analyse machine-made data click streams, mine online customer transaction data for interesting patterns of behaviour or for analysing connections and social networks for marketing, fraud detection and behaviour analysis. It is these capabilities that have made Aster Data popular with companies such as Comscore and MySpace. For example MySpace runs the database on a cluster of server nodes to help it analyse what songs and videos are going viral, what features are becoming popular, and what content is being consumed on its service.
A broadly complementary fit
These are capabilities that Teradata will want to take advantage of especially as they provide a more efficient way of querying complex and non-traditional data sources rather than its standard SQL approach. It will also help extend the company’s market reach with web-focused businesses where Aster Data has had notable success. Equally we believe Aster Data provides a natural extension to Teradata’s appliance and platform family which is aimed squarely at departmental and entry-level EDWs. Whereas Teradata has traditionally catered to the EDW needs of large enterprises through its centralised architecture, the company’s range of data warehouses are intended to lower the price points for different analytic workloads and extend the company’s reach into the mid-market. In this sense nCluster could be seen as a complement to an EDW, especially where there is a need to analyse large volumes of data from outside the EDW such as weblog files, social network data and CDR records. Equally, both companies are in a good position to forge a stronger positioning around cloud-based analytic data management platforms as both offer a range of deployment options including those for private and public clouds.
Of course there are other questions that are naturally raised by this acquisition—especially in terms of how Aster Data will integrate with other Teradata acquisitions (namely the Kickfire MySQL appliance, and recent acquisition Aprimo and its marketing analytics offering). In the case of the latter we believe there are synergies around integrating and analysing social media data, especially when driving out insights from social network graph analysis, influence mining, and so on. Similarly there are obvious business benefits for Aster Data. It can expect to leverage Teradata’s sales channels to accelerate adoption of its analytic database as well as benefiting from any increase in marketing and R&D spend as part of the global Teradata operation.
A consolidating market
The acquisition also puts Teradata’s long-term partnership with SAS under the spotlight. This corporate and technology agreement covered joint efforts on sales, marketing, services and product integration, but has come under increased scrutiny since Teradata’s Aprimo acquisition because it competes directly with SAS’s marketing automation offering. Equally, the Aster Data acquisition is likely to raise questions about the long term roadmap for developing in-database analytics for all three companies, because SAS’s capabilities were embedded in both Teradata and Aster Data’s platforms.
Similarly the acquisition comes during a busy period of consolidation in the analytics database market. In the last year SAP has bought Sybase and its columnar analytic IQ database, EMC swooped for Greenplum, IBM snapped up Netezza, HP announced plans to acquire Vertica and now Teradata has joined the party with its Aster Data acquisition. The challenge for all of these vendors will, of course, be to seamlessly integrate their acquisitions and in doing so provide an analytic database platform that can scale to meet new and expanding data volumes alongside traditional data sources as well as provide rich and deep analytic capabilities to mine all of this data. A Teradata-Aster Data combination with its embedded SQL-MapReduce approach provides a powerful punch in support of this aim.
We automatically stop accepting comments 180 days after a post is published. If you would like to know more about this subject, please contact us and we'll try to help.