Databases move closer to commoditization

EnterpriseDB got a nice $20M Series B. Congratulations & good luck to CEO Andy Astor, who’s a friend from the early days of XML and Web services.

The underlying story behind the increased traction MySQL and EnterpriseDB are seeing is that SQL databases are getting commoditized at an increasing pace. Oracle, IBM and MS may not like this but that’s where the world is headed. It’s about time–SQL has been around for 30+ years. On top of this, application frameworks have advanced to the point where some of the heavy lifting that went into databases for added functionality and scalability has moved up into the application runtime tier. Comparatively speaking, much less code is database-specific nowadays than ten years ago.

To sell databases today, you either have to show significantly reduced TCO, which is the open-source startup way, or add significant functionality. The big guys are taking the predictable fatware path–adding more features such as advanced XML processing and high-end BI for ever-smaller user audiences. That’s not a bad way to go when most of the revenue growth comes from up-selling into existing accounts.

Startups have an additional opportunity to identify large market niches and build special-purpose systems that provide 10+x advantages compared to traditional approaches. Here are some examples:

  • Doing well
    • Netezza, which goes after high-end data warehouses with complex BI queries.
  • Execution problems & small exits
    • The XML database startups: Neocore which is having a second life as Xpriori, Ipedo, … Larger XML DB/server players such as TigerLogic at RainingData and SoftwareAG’s Tamino are also facing increased pressure by the native XML features in the Big Three.
  • Big plays, too early to tell
    • StreamBase, which focuses on processing high-volume real-time information streams.
    • DataGrid, going after semantic information processing
    • Dataupia, a very cool startup by the founder of Netezza we recently invested in.

About Simeon Simeonov

Entrepreneur. Investor. Trusted advisor.
This entry was posted in startups and tagged , . Bookmark the permalink.

6 Responses to Databases move closer to commoditization

  1. Andy Astor says:

    Well, first of all, Thanks for the kind words, Sim. I am really pleased about the great reception we’ve gotten over the past 15 months in the marketplace, both in terms of customer adoption and from smart VC’s such as yourself… 😉

    I completely agree with your comments about the commoditization of the SQL database space. And it *is* about time. As you say, databases have become over-featured (mostly to protect their pricing structures), yet most customers use only a fraction of these features, and resent the lock-in from the vendors. We’ve seen this pattern before in mature, commoditized markets, where new players arrive to disrupt the status quo…Think Amdahl/IBM, MCI/AT&T, Linux/Unix, and Dell/Compaq… Frankly, open source in the enterprise software market is really about commoditizing stuff that has simply become inflexible and unaffordable.

    Always good to hear from you, Sim.

  2. Sim, any thoughts on Vertica, which was co-founded by Michael Stonebraker (co-founder of StreamBase among others)? Apparently they are in the high-end data warehouse biz, but not may details are available.

  3. Kurt, while Mike Stonebraker and I have talked in detail about StreamBase we have not talked about Vertica. Yes, they are in high-performance data warehousing but most of the info beyond that I have is second hand. I’ll ask Mike and perhaps he’ll comment here.

  4. Thanks for your question Kurt. Vertica is based on the premise that “one size does not fit all”. The market leaders (as well as the recent DBMS appliances) are all selling aging legacy software architectures, oriented toward OLTP. In other markets, especially ones that are read intensive, there is a breathtaking performance advantage from a new software orientation. It’s too early for me to share Vertica details publicly, but we will be releasing a product during the 4th quarter of 2006. In the meantime, we are doing that on a private basis with potential customers and partners.

  5. web says:

    How about a “compare and contrast” note between Streambase and Vertica?

  6. web, that’ll be comparing apples & organges ‘cos the two deal with different data types, amounts & increase rates. If you have specific questions about Streambase & Vertica, shoot.

Leave a Reply