At Microsoft's (MSFT 0.45%) Build conference for developers in May, the company announced Cosmos DB, an innovative database service. That same week, it also announced it was the first large cloud company to build data centers in Africa, further cementing its position as the No. 2 cloud company in the world.

These two announcements are exciting for Microsoft, which has done a better job building out its own cloud-services platform than any other company except cloud pioneer Amazon (AMZN 0.83%) with its Amazon Web Services (AWS). At the same time, both announcements are potential threats to the core business of Oracle (ORCL 1.88%). Here's how.

Woman with worried expression looking at laptop

Image source: Getty Images.

Database migration

One of the key features of Cosmos DB is a database migration service, which can transfer data from on-premises infrastructure to Microsoft's Azure cloud. In prior years, Oracle's SQL-based database software was industry standard, but was also very inflexible, as it had to work with "old" hard disk drive technology. That made switching from Oracle-based infrastructure a costly and risky process. This allowed Oracle to charge high prices, and make operating margins of 30%-40%.

However, recent technology advances in solid state drives, computing power, and memory have enabled the creation of less-rigid database technology called NoSQL, with leading cloud vendors offering their own NoSQL-based database software.  With Microsoft unveiling what seems to be an advanced data-migration service that makes traditional data migration easier, it is taking direct aim at Oracle's main cash cow.

Microsoft is actually the second company to offer such a migration service, following -- you guessed it -- Amazon, which unveiled its own data migration service in the fall of 2015. In a recent blog post, Amazon's Jeff Barr gave an update on the service, which as of March had migrated 20,000 databases. More frightening to Oracle should be the fact that those numbers are accelerating, with 10,000 migrations since October. Now that Microsoft has entered the fray with its own new advanced SQL database, Oracle faces a second threat, plus validation of the migration concept.

This does not mean that Oracle is doomed, but it's going to be less dependent on the stickiness of its core product to generate profits, because switching just became much easier. That means it needs to have the best database software and service on an ongoing basis. As the leader in databases for decades, Oracle may be able to do that, but having to go toe-to-toe with Amazon and Microsoft, with their full cloud and AI platforms and Amazon's tradition of undercutting competitors on margins, seems like a more difficult task.

The cloud leaders are expanding their lead

Microsoft and Amazon also have significantly more resources to boost their respective cloud offerings. IDC Research says that public cloud services should grow by 25% through 2020, and while Oracle has reluctantly made a transition to its own cloud platform, it is still a tiny player compared to the big three of Amazon, Microsoft, and Alphabet (GOOG -1.22%) (GOOGL -0.95%). Moreover, these larger companies have vast resources to aggressively build expensive data centers, further cementing their leading positions in this secular growth industry.

In the same week as Build, Microsoft announced it was going to be the first cloud company to enter Africa, with two large data centers in Johannesburg and Cape Town. In 2016, Amazon, Microsoft, and Alphabet's Google spent $31.54 billion on data centers, as the leaders attempted to deter smaller companies from hoping to compete, creating a long-term oligopoly.

Oracle's net income for the last twelve months was less than $9 billion, yet it paid out $2 billion to shareholders as dividends and buybacks, so it would have to essentially exhaust all its resources to build data centers. Even if it did, it would likely still be far behind.

Additionally, wide adoption of cloud computing is a general threat to Oracle. Before cloud, Oracle and other traditional vendors made clients predict their database capacity over the years, and would perform audits to make sure agreements were adhered to. This often forced companies to purchase more capacity than they needed -- Amazon's CTO Werner Vogels thinks companies regularly paid 30% more than necessary. Since cloud computing allows for flexible capacity -- meaning companies pay only for what they use in real time -- this practice is going the way of the dinosaur...along, potentially, with Oracle's excess profits.

Oracle counters it doesn't need the scale of an AWS or Microsoft because it optimizes its data centers for its own suite of software products to run more efficiently, and it has better tech (a debatable assertion). One can still run Oracle software in Azure or AWS, so it's not like people will no longer use Oracle's products. Still, with Amazon and Microsoft spending mightily, lowering costs, and and iterating quickly on their own enterprise offerings, Oracle will have to compete extremely hard just to maintain market share.