Oracle’s Cloud Strategy Based on High Licensing Fees and Free On-Premises Hardware: Will it Work?

Oracle has been investing heavily on its cloud technologies. I recently attended their Oracle World Conference in Delhi and came out impressed with the catchup it has done with Amazon Web Services (AWS) and Microsoft Azure.

Oracle did not have a lot of success attracting larger clients to move to its Cloud Platform. I could identify two strategic bets they have made to increase market adoption, which I will describe below. Oracle has been extending its licensing on AWS, Azure and other cloud platforms for many years now. Thus, you can provision Oracle Database on AWS Relational Database Service (RDS), and provision a VM in Azure with Oracle Database on it. You would then get license directly from Oracle for that database. It has worked well until now. A lot of our consulting clients use Oracle on AWS and Azure and were reasonably happy with this arrangement. In a very strategic move, Oracle suddenly changed the way Oracle license is counted (I will write a more technical article on it later). This resulted in 300% increase on Oracle Database licensing fees on AWS and Azure overnight. Suddenly, companies that heavily used Oracle Database on AWS & Azure, saw their bills skyrocket. This seems like a calculated strategy from Oracle. Make licensing costlier on AWS and other cloud platforms, and make it relatively cheaper on their own Public Cloud to make it look attractive. Another strategic bet Oracle made is that, it realized that many companies want to keep using Oracle on-premises, but a major reason they move to Public Cloud is the reduction of upfront hardware cost (CAPEX). Thus, Oracle started giving their physical on-premises hardware for free (or reduced cost) and instead charge customers a monthly usage fee for that hardware. This usage fee is dependent on actual usage, so if you use the installed hardware more, you pay more. If you use it less, you pay less. Thus, the pricing feels more like the pay-per-use pricing that customers are used to seeing in Public Cloud with no upfront cost. (Although, they charge a minimum usage fees, even if you don’t use the hardware at all).

This on-premises hardware also has the capability to work with resources provisioned on Oracle’s Public Cloud and thus, supports a seamless hybrid cloud scenario. Many companies might want to keep the database provisioned on-premises, while other layers of applications being provisioned on Public Cloud. This makes it easy to implement this kind of hybrid scenarios and it also helps with adoption of Oracle’s Public Cloud. Thus, Oracle is trying to reduce the need for companies to move to Public Cloud by providing cloud like billing for on-premises and making it easier to use Oracle’s Public Cloud when they do buy the hardware from Oracle. Will this strategy work? I see the below pros and cons.

Pros:

1. Many customers that are not able to afford licensing on AWS and Azure, might be forced to move to Oracle’s Public Cloud or buy their hardware with monthly billing.
2. Companies that have not moved to cloud and are considering AWS or Azure, will certainly consider Oracle’s Public Cloud or buying Oracle’s hardware with monthly billing. This might drive adoption.
3. Some companies might go with hybrid workload, where non-oracle workloads might work on AWS, and Oracle workloads might work on Oracle Cloud or Oracle on-premises hardware. There could also be co-location based solution, where Oracle DB might be deployed on Oracle Cloud and the apps accessing it might be running on AWS or Azure.
4. For companies that do not want to move the database to Public Cloud due to security or compliance reasons, the Oracle hardware in pay-per-use model will be very attractive.

This all will drive adoption of Oracle’s Public Cloud and on-premises hardware.

Cons:

1. Customers hate being arm twisted. Many might try migrating to alternatives like SQL Server, MariaDB, Azure SQL and AWS Aurora. It might drive adoption of these alternative databases that compete with Oracle Database.
2. Many companies have non-critical workloads running on Oracle Database. At least for those non-critical DBs, companies might stop using Oracle, driving down their market share.
3. Even though Oracle has made a lot of progress in its Public Cloud services, they lack the full breadth of services like AWS, Azure or even Google Cloud Platform. Thus, it will be very hard for companies to move to Oracle’s Public Cloud, in which case they might absorb the licensing cost increase, but still stick with AWS or Azure.
4. Even when Oracle is selling hardware in pay-per-use model, the other costs of on-premises hardware like maintenance, air conditioning, manpower needed, etc. are not reduced. Thus, many companies might still move to AWS or Azure cloud.

Thus overall, the strategy of using the licensing cost and pay-per-use hardware to attract more customers to Oracle Cloud might work wonderfully, or might backfire. What do you think? We at CloudThat are planning to introduce a course on Oracle’s Public Cloud. It will be released shortly and if you are interested in getting more information about it, please register your interest by clicking here.