What is Teradata

Big data

"Germany has the strictest data protection laws - all of our European customers would be immediately ready to store their data in a German data center," said Teradata's chief technologist Stephen Brobst in an interview with COMPUTERWOCHE during this year's Universe event in Hamburg. Reason enough for the data warehouse specialist (DW) to respond to the wishes of the user companies and to set up their own cloud data center on German soil. However, the IT provider does not yet want to reveal when and where exactly the Teradata data center will go into operation.

Those responsible for the group revealed a few more details about the planned service structure. The German data center is to be set up for "Managed Cloud Services". Customers who want to operate Teradata solutions in a public cloud, on the other hand, are referred to the corresponding platform from Amazon Web Services (AWS). The corresponding public cloud offer was recently activated and, according to Amazon, represented "the largest product launch of business software on the AWS platform to date". The cloud offerings are supplemented by extensions on the in-house platform. Teradata has presented "Intelliflex" for this purpose, a platform with which both internal and hybrid cloud structures can be set up.

"Very Big Data" in real time do not fit into the cloud

The various cloud offerings via AWS and Teradata's own data centers are a factor that could confuse customers with regard to the new Teradata strategy. This could also be attributed to the fact that modern big data projects, especially in the area of ​​the Internet of Things (IoT), and cloud computing do not seem to want to go together. For example, if you want to analyze sensor data from extensive IoT systems in real time, you will quickly reach the limits of cloud computing. Big Teradata customers also confirm this. "We are very interested in their new cloud offerings, but only to test new algorithms or to carry out non-time-critical follow-up analyzes - only an in-house solution is possible for production operations," said Gerhard Kress, boss of the mobility service at Siemens, which, among other things, oversees a large IoT project in the area of ​​predictive maintenance at the Spanish railway company Renfe.

Teradata management also admits that "very big data" and real-time processing are hardly compatible with cloud computing. "We see our cloud offering primarily as a sales tool. Potential customers can intensively test the performance of our platform here before they decide on a complex in-house installation," explained Dan Harrington, who is responsible for the international at Teradata Business is responsible.

Loss of sales and profits

It remains to be seen whether this will slow the current decline in sales. Teradata closed its most recent fiscal year 2015 with a clear minus. Compared to 2014, sales fell by seven percent from 2.7 to 2.5 billion dollars. And a profit of $ 367 million at the time turned into a hefty loss of $ 124 million. This is also due to the exit from the market for marketing software announced in the third quarter of last year. But the bad numbers are primarily due to the fact that the classic DW area no longer needs high-wire systems. Many tasks that once required a high-priced Teradata solution can now be handled using cheaper cloud offerings.

The data warehouse market is only growing slowly

The market researchers from Market Research Media are forecasting only meager growth of 8.3 percent over the years for the entire DW market in the next five years. With the simple DW solutions, their market researchers see an increasing trend towards the use of "DW-as-a-Service". Gartner also sees massive restructuring in this market. "Within the next two years there will be dramatic changes. The market share of traditional DW providers will decline, while at the same time new big data platforms such as Hadoop or NoSQL clickstream data will gain in importance," it says in the latest Gartner report on the DW market.

In summary, this means that the traditional DW market is practically disappearing. Depending on the data and performance requirements, these solutions would either be replaced by cloud applications or by inexpensive on-premise standard systems. This also includes the new open source platforms such as Hadoop. For high-performance systems, however, from the analysts' point of view, only the area of ​​"extreme big data" remains in a real - or at least approximate - real-time environment.

There is room for improvement in advanced analytics

Teradata's strategy is to be represented in all segments as much as possible. However, a leading role is not realistic for Teradata - at least in the lower areas. The competition here is too great and the margins are too low. In contrast, the specialist has good chances in the upper performance classes, but here the market still has to be developed strongly. "Many of our customers believe that advanced analytics in connection with the immense amount of data could lead to important insights and business processes, but they don't know how to approach something like that," says Ron Bodkin, president of the Teradata subsidiary Think Big, a pure company Consulting company that specializes in big data analytics, the current situation. In many companies he sees a serious discrepancy between data acquisition on the one hand and the analytical possibilities on the other. "All large companies are now shoveling up immense heaps of data, but they lack the know-how to evaluate this data in a targeted manner," he says from his everyday work.

In addition to Think Big, Teradata also operates other consulting facilities. The provider recently set up a global "IoT Analytics Team" that is supposed to offer special assistance in the analysis of big data in the IoT environment. The team is part of the Teradata Lab and has offices in the US, UK and India. However, not all Teradata customers rely on outside advice. In addition to the Siemens team mentioned above, the car manufacturer Volvo already has the necessary analysis know-how. A team of around 30 under the direction of Jan Wassen works primarily on solutions in the area of ​​predictive maintenance. "We can already answer two important questions very well: What happened and why did it happen - now the next big question is, namely 'What WILL happen'", is how Wassen describes his area of ​​responsibility.

It's all a question of culture and organization

Just as once with the introduction of BI or even further back with the emergence of ERP, the problems are less of a technical or mathematical nature, but rather fall into the categories of corporate culture, lack of mutual understanding or organizational inhibitions. "You have to have a lot of staying power," comments the Volvo manager, somewhat resigned to the day-to-day problems of his staff department.