Let’s get this straight: Google and Microsoft are both growing their respective cloud businesses roughly 100% every quarter, each on a reasonably solid base. Amazon We Services (AWS), while not growing as fast, isn’t expected to do so, given that it’s now on a $20 billion run rate. As such, 45% growth is not merely respectable, it’s amazing.

Oracle’s PaaS and IaaS cloud revenue growth, by contrast, barely beat AWS’ at 49%, yet on a much smaller base of under $416 million. Unfortunately for Oracle, it’s not bound to get much better anytime soon, with the company projecting a top end of 21% growth for next quarter. The reason? Oracle simply can’t get a developer clue, a fact that will continue to haunt its earnings for the foreseeable future.

Raining on Oracle’s cloud parade

Oracle tried to dress up its cloud earnings as well as possible, by progressively narrowing the definition of cloud to get higher growth numbers. For example, the company reported total PaaS and IaaS revenue at $416 million, up 24%. Given how limp that growth rate is, the company then noted that a big chunk of its SaaS and PaaS revenue is actually comprised of its legacy hosting business. Take that out and boom! The growth rate jumps to 49%.

In other words, the company wants to pad its total cloud revenue number with non-cloudy legacy hosting, but when it wants to show revenue growth, it needs to trim that revenue out. Even so, 49% is not competitive with the big cloud vendors.

SEE: Everything as a Service: Why companies are making the switch to SaaS, IaaS, PaaS, and more (Tech Pro Research)

Worse, if we take Oracle’s overall cloud growth, including its relatively strong SaaS business, we see a troubling pattern of declining growth rates. Oracle is projecting its overall cloud business to grow by a maximum of 23% in Q4 2018. For those keeping track at home, that’s down from Q3 2018’s 32% growth rate, which is down from Q2 2018 growth of 44%, which is down from Q1 2018 growth of 51%, which is down from Q4 2017 growth of 58%. Oof!

Not so fast, co-CEO Mark Hurd was quick to add on the earnings call. If you take an even smaller slice of the company’s cloud revenue–public IaaS–the growth rate jumps to 142%. Of course, that growth is on top of a comparatively tiny base, but Oracle wouldn’t divulge that revenue number. In their shoes, neither would I.

Oh developer, where art thou?

Chairman Larry Ellison then hopped into the fray, declaring that: “No other cloud provider has anything like [Oracle’s fully automated database].” That sounds cool, except according to Stack Overflow developer survey data, no other cloud company needs to rush to build one, either.

Despite Oracle’s traditional heft in databases, it can barely crack the top-10 in developer affections. Oracle’s eponymous database comes in at no. 9 in terms of developers’ preferred databases, with 11.1% listing it. The company’s MySQL acquisition, however, sits on top of the developer heap. If only Oracle were more interested in a fully automated MySQL….

Ask developers what databases they love most, and Oracle barely squeaks into the top-20 (in the twentieth spot). Amazon, Microsoft, Google, MongoDB, and other vendors all make the top-10. One area that Oracle does shine, however, is in the list of “Most Dreaded” databases: Oracle comes in second, trumped in developer animus only by IBM’s DB2.

SEE: Special report: The art of the hybrid cloud (free PDF) (TechRepublic)

Ask those same developers which platforms they’re building on, and AWS, Google Cloud Platform, and Microsoft Azure all make the list. Oracle? Nope.

This ongoing lack of developer appeal will continue to haunt Oracle, and it’s unclear how the company can turn it around. Certainly the company’s continued emphasis on software licenses don’t help–on its earnings call, Oracle mentioned software licenses 53 times. Cloud did win out with 71 mentions, but for a company earnestly trying to reshape itself, this continued focus on its old business isn’t helping it to win new developer affections.

Oracle is actively trying to hire a new executive to run developer relations. This can’t happen soon enough, but it also likely won’t be enough. To reach developers, Oracle needs to be a completely different company, one that delivers copious quantities of services at minimal entry cost. That’s the exact opposite of its legacy business, which is hampering the company’s efforts to change.