Greenplum, a petabyte-scale database for data warehousing and business intelligence on Monday 6/8 made three separate major announcements:
- Greenplum Brings the Power of Self-Service to Data Warehousing with its Enterprise Data Cloud™ Initiative
- Greenplum Announces Greenplum Database 3.3
- Future Group Sets the Standard for Retail Analytics with Greenplum Database
The major announcement out of the three is the Enterprise Data Cloud (EDC) positioning. Much has been written by the press and analysts around the capabilities and technology related to the announcement. My focus in this post is to analyze their positioning and messaging that was used with the launch and, by way of contrast with a competitor Aster Data, to offer some thoughts around the marketplace and marketing so far in the high end analytic database space.
For this exercise, let’s use the press release itself, Greenplum Brings the Power of Self-Service to Data Warehousing with its Enterprise Data Cloud™ Initiative, as the spring board. The core messaging stated in the release around the platform and methodology is clear and precise. Although the announcement could be construed as “me cloud too”, there is enough of a spin around self-service data marts and the associated benefits to emphasize that this is not just putting Greenplum’s platform into the cloud as is. Instead, the message is mainly targeted at the business user and not so much IT. Saying that the business user can quickly spin up data marts without having to go through IT speaks much more to the some of the cloud computing business benefits, rather than just pure technology efficiencies. However, to be clear, Greenplum’s customers are not running on public clouds such as Amazon EC2, rather they are running “private clouds“. Greenplum says that it hasn’t yet seen a demand for public clouds, yet Aster Data a Greenplum competitor has public cloud references.
Last night, I attended Scale Camp in Santa Clara, CA where many technical developers presented architectures of their implementations, and shared experiences of using technologies such as Hadoop, MapReduce, Amazon Elastic MapReduce, Cascading, Bixo, Katta and more. Among all the bits and byte presentations, I narrowed in on a significant comment by Paco Nathan of Sharethis.com. He was presenting his company’s mashup of technologies for the cloud for big data analysis, which included many of the technologies I mentioned, and Aster Data’s ncluster cloud edition. Sharethis is a customer of Aster Data, and as I mentioned, Aster Data is a competitor to Greenplum and offers similar technology, both have Postgres, massively parallel processing (MPP) system and support SQL and MapReduce as a key element of their architectures. Unlike Greenplum however, Aster Data already has references running on public clouds.
Aster Data’s announcement on 6/9 around their .NET support further emphasizes that much of the focus today is on feature and function positioning to the IT developer/buyer. However, Merv Adrian, a leading independent analyst points out that “it will drive a relatively unopposed business opportunity for a while in a specific underserved (.NET) community.” In contrast, Greenplum could be trying to move “up the stack” and keying in on the business buyer, and potentially the ultimate decision maker, since more and more companies are requiring direct business benefit justification for procurement of technology. Aster Data in contrast, also has a very slick web admin interface to control and manage the deployment of instances of their database, which I saw demoed at SDForum. Given this I wouldn’t blame Aster Data if they were quietly simmering right now knowing that technology wise they were, and may still be way ahead of Greenplum architecturally, but Greenplum’s release has managed to position themselves as though they have come up with the latest and greatest thing to hit the cloud since Google App Engine. Although details are sparse around the promise of Greenplum’s future admin/business facing web console (coming in version 3.4), they could put more of “a face” onto technology that a business user cannot see or touch today. If so the commodity battle of who has the faster or more integrated infrastructure will no longer be the main battle ground, rather with the cloud, the business user experience and their self-service ability could be a major factor and a key differentiator in deals. I’m not saying that any of these vendors would ever offer a slick query interface that is the domain of the BI vendors like Microstrategy. The question is will all high performance analytic database vendors look to deliver more capabilities which are focused on business users, or will they stay within the realm of the IT buyer, and focus on the bits and bytes of the management, integration and administration of the database. Greenplum is certainly positioning themselves in that direction with their self-service data marts announcement.
The ecosystem section of the press release is also quite telling in that it contains quotes from a variety of partners. I applaud the effort taken to gather, gain approval and coordinate all of the quotes within the release itself. More importantly, it shows that Greenplum has active partnerships which are also potential channels for go to market. Although Greenplum can run on commodity hardware, they invested in deploying on Sun (back in 2007) and EMC (announced 12/08) in order to build enterprise credibility, but also no doubt as a means for having another major sales channel beyond direct sales. Additionally, they added quotes from Informatica (to subtly reflect that they have partnerships for getting data into the cloud) and GoldenGate Software (disclosure, one of my former companies) for high availability and real-time change data capture.
To close, I should just mention that there are many other excellent players in this crowded space of high performance analytic database vendors. For example, Vertica a columnar database made a new release announcement for their Vertica On-Demand service (which they have had since 5/08). They too have public cloud references. ParAccel, another player has had hardware vendor relationships with both Sun and more recently also EMC. As far as I know, they haven’t done anything with the cloud yet.
There’s no doubt that the technology in this space is maturing with certain features and functions fast becoming commoditzed and the positioning and messaging is sounding very same-ish. If I were to take a position, I’d give Aster Data the lead on technology and architecture, with public cloud references a good 4 months ahead of Greenplum’s announcement and their recent announcement of .NET support, while I give Greenplum kudos for not avoiding a “me cloud too” label with business focused positioning for their EDC initiative. Even if it does somewhat cover up the fact that they might a little late to the cloud party and they are purely deploying on private clouds.
There are plenty of opportunties for all, but now would be a good time to for all these vendors to up their product management and marketing game, to get to the heart of what will truly differentiate and stand out in this growing market.
Nice analysis and very much on point. (Disclosure of my own: Greenplum is a client.) A couple of points of interest for me:
1. there are tons of opportunities out there for this kind of solution, and several players (you named them) can all grow succesfully on the scale they are at right now – without going head to head all that often.
2. there are differences in positioning, and you got to a key one between Aster and Greenplum – the .NET announcement from the former is, IMHO, very big for them, and if it takes will drive a relatively unopposed business opportunity for a while in a specific underserved community. If they execute on it, it ought to be a nice bump in their business.
3. At the same time, Greenplum is indeed going upmarket with their story, selling an effectiveness and especially cost-effectiveness story about easy resource allocation, build-up, and tear-down as needed to people other than the ones doing the coding – who are willing to pay for agility. For both, it’s also a big opportunity vis-a-vis the big expensive proprietary players.
4. I don’t see much evidence of dramatic technical differentiation yet (other than the .NET thing) that we can get at, until a few more implementations from each are out there to be compared. Both (and some others you mention) have success stories to tell. And at these prices, success in the next year or two may have as much to do with building out the sales, marketing, and support infrastructures as anything else.
The good news – this is more fun than we’ve had in the database market for years!
Thanks Merv. Great comments. You are correct that .NET is underserved as a developer market, I neglected to point that out, even though I was thinking it at the time I read the announcement. Got a little too myopic about biz vs dev audiences. I amended the post to include your quote 🙂 And I agree, the market is ripe with opportunities for these new technologies for handling big data!
Really interesting take – IMHO GP is playing the “intellectual leadership” card while Aster tends to play more on the “gear-head” level. Dotnet is underserved in the sense science/research/financials pretty much belong to the ‘nix/LAMP world I believe. In either case if I had to bet on LOB users vs. researcher/quant/techies for the future of BI I think the LOBs have it 🙂 – I dont think too many users want to depend on PhDs to run their BI. I could be way off on that.
Thanks for the excellent post!
Ramon, appreciate the insightful post. Trying to get my head around all the cloud hype. outside of the fact that a database is accessed via an IP-address and that IP-address can reside anywhere from any tool. I’m getting the impression that the value proposition of “cloud enabled” is around
1) the ability to subscribe to the services and avoid the costs of captial investment (public cloud model) or subscribe to the services and leverage a shared-service capital investment (private cloud model) and
2) the ease-of-admin/ease to stand-up/tear-down with admin gui’s thus the play down of needing IT to do it.
I’m reading between the lines. Curious on your take.
Hi Mike, thanks for the comment. I would concur with the two points you list above as main benefits of leveraging the Cloud. I would also add that apart from cost and ease-of-use, the Cloud presents an opportunity to deliver and deploy systems and applications/functionality and resources at “the speed of business”. Numbered are the days of business users having to wait for IT to get around to providing capabilities, resources and functionality to support ideas, go-to-market goals and strategy. Those companies that embrace the cloud will find themselves leaping ahead of the competition IMHO.