Life used to be so simple for Oracle.

Back in the good ol’ days, the company could build a great database, charge a king’s ransom for it, and milk the maintenance stream forever. Rinse. Repeat. Unfortunately, this model no longer works, as a series of earnings misses over the last few years have shown.

While a number of factors are at play in Oracle’s stumbles, one of the most persistent is the rise of open-source databases, both relational and non-relational (NoSQL), as a recent Bloomberg article posits. As Powa Technologies CEO says, “They scale and operate extremely well, and they don’t cost anything.”

This sounds like a recipe for yet another Oracle earnings miss. But it may not bode well for the open-source database companies, either.

The stumbling giant

Oracle’s struggles to adapt to the future have been well-documented. As detailed by The Wall Street Journal a few months back:

“Oracle can no longer count on its near-monopoly in database systems, as the emerging technologies of big data and cloud computing–often available in open-source versions that cost far less to use–along with a customer base eager for alternatives, fragment the market. At the same time, younger competitors such as and Workday are picking off Oracle customers by offering specialized software applications sold by subscription rather than in a large lump sum plus a service contract.”

To its credit, Oracle has been juicing its cloud business to ward off the decline in new license revenue, which has declined for seven straight quarters. This is helping it to hedge against the Workdays of the world, not to mention the general shift toward subscription-based business models.

But what it’s not doing, as the Bloomberg article describes, is protect the database giant against the open-source pygmies. Indeed, as Don Duet, co-head of technology at Goldman Sachs, insists, “It’s hard not to go into our datacenter and see a tremendous amount of open source running our applications and middleware.”

While open-source databases are finding their way into large organizations, it’s really their hold on tomorrow’s Fortune 500 companies that should be of most concern to Oracle. Surveying startups valued at $1 billion or more, Bloomberg found that “None of the companies surveyed indicated they had a large Oracle database deployment for their main services, though many used bits of Oracle software to run aspects of their organizations.”

The inverse, in other words, of the open-source databases’ role among today’s Fortune 500.

That’s music to the ears of the companies providing that software, like MongoDB and DataStax (Cassandra), but it can’t sound sweet to Oracle chairman Larry Ellison.

Sucking the life from the market

Before the open-source database startups uncork the champagne, it’s worth remembering that “free” doesn’t pay anyone’s bills. It might be nice to empty Oracle’s pockets, but currently, the money lost from the market isn’t being recouped by the open-source upstarts.

Indeed, former MySQL CEO Marten Mickos once quipped that his goal was to “shrink [a $9 billion relational database market] to $3 billion and take a third of the market.”

That’s great. but it’s also deflationary. While good for database users, it’s not necessarily a boon to database vendors.

Wikibon analyst Jeff Kelly highlights this quandary in an excellent article, “The Challenges of Building A Thriving NoSQL Start-up.” As he writes, “There is no question that the NoSQL database category is on the right side of several secular trends…, [b]ut the challenge for…open source NoSQL database companies is to turn open-source users into paying customers in large numbers.”

To be clear, companies like MongoDB and DataStax do make significant amounts of money, but nowhere near the value being sucked out of the market. And, as Kelly suggests, each has dramatically fewer customers than users, which means that without “massive scale, the compan[ies] end[] up taking in less revenue than the value of [their] software could command.”

In short, both can credibly claim millions of downloads and even tens of thousands of users (in the case of MongoDB). But paying customers? That’s harder.

“Harder” doesn’t translate to “impossible,” and I spent enough time at MongoDB to believe that open-source databases have a bright future ahead of them. Even so, it does mean that the industry is still coming to grips with the popularity of open-source alternatives to Oracle’s hitherto impregnable position atop the database heap. It’s not as simple as “Oracle’s loss is open source’s gain.”