Database market set for seismic shift driven by accelerated cloud and SaaS adoption

451 Research has been ruminating about the potential impact of the cloud on the database market for several years, informed by data generated by our Market Monitor revenue and market size estimates, as well as customer insight from our Voice of the Enterprise (VotE) service. Combining data from both provides a projection that indicates a seismic shift in the database market between now and 2021.

The 451 Take

For several years, we have been describing the data and analytics sector as being on the brink of change, but the implications of that change, in terms of evolving mainstream adoption trends, have seemed a long way off. The combination of data from 451 Research's Total Data Market Monitor and VotE: Cloud illustrates that the database market is accelerating toward change, at least in terms of deployment and execution venues, as revenue from on-premises non-cloud deployments declines rapidly between now and 2021. It would be easy to assume that this shift will primarily impact the database incumbents. However, with revenue to be shared across multiple venues (on-premises non-cloud, on-premises private cloud, IaaS and SaaS/PaaS), it will be those vendors that offer a consistent experience across the broadest portfolio of options that will likely benefit most from the changing market dynamics.

The potential impact of the cloud on the database market has been the focus of multiple reports from 451 Research over recent years, with a variety of conclusions and advice. For example, back in 2013, we noted that database as a service (DBaaS) was one of the driving forces encouraging adoption of next-generation databases, before quickly following up with a warning that database startups struggling to gain adoption in the on-premises database market shouldn't, however, expect an easy ride on the cloud.

The following year, we noted that the fact that the majority of enterprise database workloads are deployed on-premises might mean that private clouds, rather than public clouds, could be key to greater DBaaS adoption. More recently, in 2016, we explained the drivers and challenges of public cloud DBaaS adoption, and argued that there would not be a seismic shift toward IaaS and SaaS within the next two years.

However, the latest data from 451 Research's Market Monitor revenue and market size estimates, combined with customer insight from our VotE service, indicates that a substantial shift in the database market is now on the horizon.

Multiple trends

One of the reasons we've presented multiple (although we would argue not contradictory) views on the database market is that there are various trends at play, and these are impacting database workloads in different ways.

For example, while the majority of current database workloads are deployed on-premises and few are being moved to public clouds, we do see greater adoption of public clouds for new database-based workloads. This is especially true of non-mission-critical workloads – such as for development and test, or backup and recovery purposes.

Meanwhile, we also believe there is a greater tendency for non-transactional database workloads to be spun up in the cloud, which would suggest more rapid adoption of DBaaS for non-relational (i.e., NoSQL) databases than traditional relational databases. However, relational databases continue to drive the bulk of overall database revenue, and are therefore the primary focus for both cloud providers and incumbent relational database vendors in trying to drive greater DBaaS adoption.

Then, of course, it is also important to note that DBaaS could relate to both private and public cloud database services, while on-premises database revenue could include traditional non-cloud deployments, as well as on-premises private clouds. Meanwhile, public cloud database revenue includes both database software deployed on public IaaS, as well as consumed via packaged PaaS and SaaS offerings. There are also off-premises non-cloud, and hosted private cloud, options.

The result, if you're looking for simple answers related to database adoptions trends, is that there aren't any. It is a riddle, wrapped in a mystery, inside an enigma.

From a 451 Research perspective, another reason for presenting multiple views of the database market is that we have multiple indicators – including Total Data Market Monitor, which provides five-year forecasts of overall database revenue trends, as well as VotE, which provides details of customer adoption trends across the various consumption models, but only projected for two years.

By building data from VotE (specifically, Voice of the Enterprise: Cloud Transformation, Workloads and Key Projects) into the models that underpin the latest update to Total Data Market Monitor, we can get a better picture of the trends we see impacting the database market in terms of execution venues over the next five years.

The combined result is perhaps best thought of as an informal projection rather than a formal 451 Research revenue estimate (since the combination isn't subject to the usual rigorous Market Monitor research methodology). For that reason, we are only providing segment share figures rather than dollar amounts. However, those segment share figures are worth discussing in further detail since they indicate that a seismic shift in the database market is on the horizon.

The stats

The combined Total Data Market Monitor and VotE data illustrates a point made above – that the bulk of database revenue today is generated from on-premises, non-cloud deployments. These represented more than 72.7% of database revenue (both operational and analytic databases) in 2016.

Combined operational and analytic database revenue is expected to increase at a compound annual growth rate of 8.3% between 2016 and 2021. However, the proportion of that revenue generated from on-premises, non-cloud deployments is expected to decline rapidly between now and 2021, by which time it will represent just 22.4% of combined database revenue.

The proportion of database revenue generated from off-premises, non-cloud deployments (essentially databases hosted on traditional infrastructure) is also expected to decline, from 6.2% in 2016 to 5.2% in 2021.

While those figures indicate a rapid rise in cloud-based database deployments, it isn't a simple matter of on-premises non-cloud deployments moving to the public cloud. There are a variety of options for cloud-based database consumption, most of which are expected to grow in the coming years.

The exception is hosted private cloud (which involves a database hosted by a provider using private cloud infrastructure), which represented 5.4% of database revenue in 2016, and will decline to just 3.4% in 2021.

In comparison, on-premises private clouds – which would include both managed or self-operated database software installed on third-party private cloud infrastructure software (OpenStack) as well as commercial on-premises 'cloud stacks' (Microsoft Azure Stack and Oracle Cloud at Customer) – represented 9.9% of total database revenue in 2016, but that is expected to increase substantially to 21.1% of total database revenue in 2021.

The biggest growth is projected to come from public clouds, however. IaaS/public clouds represent revenue generated by database software vendors driven by a user installing the database software on public cloud IaaS. It accounted for just 3.3% of total database revenue in 2016, but is expected to climb to 26.3% of total revenue in 2021.

Additionally, SaaS/PaaS represents revenue generated by database software vendors driven by a user consuming a DBaaS offering (which for the purposes of this segmentation includes PaaS, and may or may not be running on IaaS/public clouds). The proportion of database revenue generated from SaaS/PaaS was only 2.6% in 2016, but is expected to climb to 21.6% in 2021.

As a result of all this, the proportion of revenue coming from non-cloud environments is expected to decline from 78.9% in 2016 to just 27.6% in 2021, while private cloud options will increase from 15.3% of total database revenue in 2016 to 24.5% in 2021, and public cloud options will increase from just 5.9% of total revenue in 2016 to 47.9% of all database revenue in 2021.

Meanwhile, combined on-premises database revenue (both non-cloud and private cloud) is expected to decline from 82.5% of revenue in 2016 to 43.5% in 2021, while combined off-premises database revenue is expected to increase from 17.5% in 2016 to 56.5% in 2021.


The implications of all of this are pretty clear. Both database users and databases vendors need to prepare, if they haven't already, for a future of multi-venue database deployments, with on-premises non-cloud, on-premises private cloud, IaaS and SaaS all projected to represent more than 20% of total database revenue in 2021.

For enterprises, this will mean an abundance of choice, which is a good thing, but also plenty of opportunities for confusion, and increasingly challenging data management requirements. It is for this reason that we expect the ability to manage data wherever it resides (from the datacenter to multiple clouds) to emerge as a key battleground in 2018 and beyond.

We will discuss more about the implications of multi-venue data management in future reports, but the ability to identify and manage data across multiple execution venues is clearly going to need to be a priority for any data-centric businesses (which increasingly means every business) for both efficiency and regulatory reasons.

The need for enterprises to understand what data they have at their disposal, and where it resides across the database estate, is also likely to drive merger and acquisition activity among data management vendors as they look to plug functionality gaps and extend the value of their existing customer relationships.

For database vendors, the charts above also illustrate the importance of offering a variety of database consumption options, particularly on-premises non-cloud, on-premises private cloud, IaaS and SaaS/PaaS. With each of these projected to represent 20-25% of total database revenue in 2021, no vendor can afford to ignore any of these options in particular – and providing consistency of experience and cross-location data management functionality will be increasingly important functional requirements.

New Alert Set

"My Alert"

Failed to Set Alert

"My Alert"