Cloud remains a small percentage of IT spending, but its gravitation pull is huge

Even though most IT dollars still go to non-cloud vendors, all the momentum is toward the big three cloud vendors: AWS, Microsoft, and Google.

istock-510476540.jpg

Image: BsWei, Getty Images/iStockphoto

It's hard to compete with cloud. Yes, I know Gartner's estimate of global IT spending in 2018 ($3.7 trillion), of which just $175.8 billion stems from public cloud services. Cloud is a mere rounding error in the overall scheme of IT spending.

And yet it's hard to compete with cloud. More specifically, it's hard to compete with the top three cloud vendors: AWS, Microsoft, and Google.

SEE: Vendor comparison: Microsoft Azure, Amazon AWS, and Google Cloud (Tech Pro Research)

All the cool kids are doing it

Enterprises must keep the lights on for existing workloads, but even these are actively being considered for cloud. Take The New York Times, which just moved an old, COBOL-based mainframe application (managing daily home newspaper delivery) to a Java-based on-premises application, and then to AWS. If the COBOL kids are getting nudged into public clouds like AWS, everything is possible.

As such, despite public cloud computing's relatively small piece of global IT spending today, its gravitational pull promises a much greater share of that spending tomorrow.

Take things like streaming data. As Lawrence Hecht has written, "Enterprises are adopting commercial stream processing offerings from both their cloud providers and more specialized vendors." He's being generous by including "more specialized vendors," because they're quickly getting displaced. Though Lightbend is holding its own (especially in a survey it sponsored, I believe), it's telling that the big three cloud vendors occupy three of the top-five spots in a survey from The New Stack. For modern, streaming data applications, the big three dominate.

It's hard to compete with cloud.

SEE: Prepare for serverless computing (ZDNet special report) | Download the PDF version (TechRepublic)

Selling what customers want

Open source is another area where the big clouds generally trump the communities responsible for the software itself. In today's market, enterprises don't really want to buy software--they want services. In such a world, the ability to operate software is worth much more than the software itself.

Redmonk analyst Stephen O'Grady nails this point:

What many open source projects have struggled to grasp, however, is that just as open source won in many settings not because it was better but because it was more convenient than proprietary alternatives, so too are cloud-based managed services more convenient in many cases than commercially supported open source offerings.

The obvious implication, therefore, is that commercial open source organizations need to compete not purely on an engineering basis, but on the axis of convenience as well. Which effectively means that they need managed service versions of their products.

Cloud services, in other words, are out-conveniencing the uber-convenient open source. You can be mad about this all you want, and invent all sorts of license gymnastics to try to block the public clouds from misappropriating "your" code (a weird thought in open source land), but developers are actually the arbiters here, and they're voting for convenient, accessible software. If the open source communities/companies don't provide it, the cloud companies certainly will. But still, I get it…

It's hard to compete with cloud.

Also see