The Teradata Turnaround
I was at the Teradata Influencer Summit last week, as you might be able to deduce from the shot of me with Angela Schmidt, one of Teradata’s impressive analyst relations team. If you’ve been following the Teradata news, you noted that Teradata appointed a new CEO in May. The new broom is Victor Lund, replacing Mike Keohler, who presided over Teradata for the previous nine years. Teradata stock soared 13% on the announcement, although it coincided with reasonable first-quarter figures which may also have influenced the market.
Lund is an experienced turnaround expert, and Teradata looks to me like a tanker that is in the process of turning around. We shall see. The bind that Teradata and several other companies (particularly Oracle and IBM) find themselves in has several aspects that can be described in a few words:
- Open Source
- Big Data Competition
Let’s take them one by one and I’ll elaborate.
The Cloud Conundrum
Almost everyone will move to the cloud. It’s just a matter of time, although it may take a good deal of time. Nevertheless, there’s already a multitude of Big Data instances in the cloud and big databases too – Amazon Redshift and Google’s BigQuery, for example. Data that might have sped in the direction of the Teradata database is going to live elsewhere.
Teradata’s solution is to put its engine in the cloud – on AWS, on Azure, and on any cloud service that can accommodate it. So you thought that the Teradata MPP was built to the hardware it was delivered on, and you were right. But, needs-must-when-the-devil-drives, it has been ported to commodity hardware.
The Open Source Challenge
All of the major corporate IT vendors have been wrong-footed by the Apache open source movement that began quietly with Hadoop but marched forward noisily with Hive, HBase, Pig, Zookeeper, YARN, Mesos, Spark and a cast of thousands of developers. The free-software-pay-only-for-support op-ex business model sharply undermines the once dominant cap-ex business model of vendors like Oracle and Teradata. It wouldn’t matter so much if customers were drowning in all the open source data lakes that have flooded the IT landscape – and by the way, some of them are – but unfortunately data lakes have virtues, too. A data lake truly is a very inexpensive place to store data, especially data that doesn’t belong in a data warehouse. The data lake tide cometh in and will not be stemmed.
So what is Teradata to do? Teradata’s response to this challenge has been to embrace and integrate. Its integration (with Hadoop and its fellow animals) is based on Aster Data. It’s practical, and it falls within Teradata’s single data layer vision. But Teradata has stepped beyond that, on the one hand embracing and extending Kafka with its excellent Listener product, and on the other becoming the support service for Presto, a SQL capability that is likely to prove increasingly important with time and may well dominate. I’ll say more about this too, in another blog post.
The Big Data Game
The truth is that no one outside of the geek community cares much for data lakes or databases or data warehouses. The boys with the budgets care about applications, and in the Big Data universe that means they care about analytics – and because they care about analytics, they will eventually care about application performance if it turns out to be awful. But in the meantime, the whole of the Big Data space is crowded with vendors pushing one or another analytics offering at you.
Now, in my opinion, this is not going to end well. The reason I think this is that I’m seeing way too much purchasing of point solutions and way too little strategic thinking. Part of the problem is this: many businesses are quite new to analytics and don’t yet understand that its adoption is a business process before it is a technology process.
Teradata also has this covered. At the summit, I was completely surprised by the power of Teradata’s analytics weaponry. It’s not without gaps; it doesn’t have graph analytics perfectly squared away yet, for example, although it can deliver some capability in that area. What it can demonstrate is multi-genre analytics and capabilities in many areas: stats, machine learning, time series analytics and so on. It can also bring consultancy skills to the party thanks to its acquisition (in 2014 – doesn’t time fly) of Think Big.
So here’s the skinny: new CEO, good strategic position, interesting tactical moves, has grasped all the nettles. Teradata looked better than I expected, and to be honest, most of what it can demonstrate was in preparation long before the new CEO took his seat. No doubt he has an interesting job ahead of him.