We just published an in-depth market landscape report called Beyond RDBMS: Not Only SQL Challenges the Status Quo. While the full report is only available to private customers, here is an infographic that highlights some of the key findings about the nature of the Not Only SQL database market.
The primary key finding, however, is that the vendors of the Not Only SQL database community are indeed making healthy progress in penetrating the market, or perhaps creating a new market. What is the best metric to support that finding? The most-often posited question I have heard when discussing the Not Only SQL database market is "how big is it?" We did indeed attempt to take on that question during the research phase associated with the just-published report.
We felt number of paying customers, number of database instances, number of free downloads, and pricing information, combined, would give us a good sense of the depth of market penetration. We obtained information from 23 commercial Not Only SQL database providers (all listed in the aforementioned infographic), and found, somewhat to our surprise, that several of the Not Only SQL database vendors had progressed nicely in terms of paid customer counts.
While we cannot share specific numbers, over five cited over 500 paying customers (and a few had reached or were north of a thousand), and over ten cited over 250 paying customers each. We estimate that there about 60,000 production database instances (many paying customers end up buying more than a single database instance) implemented by paying customers across the 23 Not Only SQL database offerings.
Sorry, make that 22 database offerings, because an important footnote is this 60,000 number does not include counts from Amazon Web Services DynamoDB which may have the highest number of instances of any of the Not Only SQL providers. Counting the number of cloud database instances for such a flexible service is daunting, and we are not yet comfortable making an educated guess.
Pricing varies somewhat wildly, but if we calculate a weighted average price per year and eliminate some outliers (note that annual subscription pricing was by far the most used pricing technique—and again the cloud service databases seriously complicate this calculation), multiply it by 60,000, and put in some plug factors for cloud services databases, and we come up with about a half billion dollars of revenue.
We know, however, that figure is inaccurate, because some of the databases being sold are true price outliers on the high side. We also have not included concomitant professional services, servers, storage, networking, and related software being dragged through by Not Only SQL databases. I prefer to think about markets the way IBM does—all in—since a service dollar counts every bit as much as a database license number counts. Add in those factors and it is clear that Not Only SQL databases, all in, will easily account for a billion dollars of revenue in 2012, and the combined providers will easily eclipse that number in 2013. We are clearly not alone on being bullish about Not Only SQL databases, for at least $300m in VC backing, and the number may be closer to a half billion because some or the VC-backed vendors did not disclose their numbers to us, are betting on changing database market.
Do the database leaders such as Oracle, IBM, and Microsoft care? I mean, what is a billion dollars to them, split three ways? On the surface, not much, but there are simply too many cases in the IT industry where leaders reacted slowly to a rising tide (Sun and Linux, Nokia, and Smartphones come to mind) and suffered the consequences. Databases are notoriously sticky; once you have one installed, have applications running on them, and DBAs trained to support them, they are very difficult to displace. It works both ways though, that mere billion dollars or somewhat more of lost revenue will not easily return to the database leaders either. I am certain the leaders care, to varying degrees, and are already watching DynamoDB, 10gen MongoDB, and SAP Hana, as well as checking in often on Teradata Aster and EMC Greenplum and anything Google is doing on the database front.
I frankly suggest they pay close attention to the entire Not Only SQL database group however, because all these new databases have emerged in response to evolving customers requirements. Though the subscription pricing models, the ease of try (or even truly use) before you buy, the performance, and scalability enhancements all contribute to Not Only SQL databases penetrating enterprises, it is really about enterprises developing a new generation of applications, and being open to new data management alternatives for those applications. The Not Only SQL movement presents established enterprise database leaders with the classic double-edged sword of threat and opportunity.