We’ve said before that a disruptive innovation needs time to diffuse throughout an economy before it can take a dominant position as an economic driver.
Consider the cable TV industry. Cable is not new but it has taken about four decades to reach its prominent position. During that time, cable providers have spent much time and resources building out their infrastructures, i.e. diffusing their products. When they weren’t building they were buying competitors and gaining market share that way.
Today we’re witnessing a similar diffusion event in building the information utility that we’ll all take for granted at mid-century. It’s not simply SaaS or even cloud computing, it’s a full-on utility similar to cable, electricity, or various forms of broadcasting.
Enterprise software vendors are working full tilt to make the vision of an information utility real and there’s still doubt which vendors will be left standing. One that has a great shot at being present at the finish line is Oracle and it’s interesting that so many financial analysts question its cloud growth numbers.
The financial reports for Oracle’s second quarter showed lumpy progress:
- Cloud Software as a Service (SaaS) revenues were up 55 percent to US$1.1 billion.
- Cloud Platform as a Service (PaaS) plus Infrastructure as a Service (IaaS) revenues were up 21 percent to $396 million.
- Total cloud revenues were up 44 percent to $1.5 billion
The disproportionate revenue tilt toward SaaS was a worrying sign to some. A “mere” $396 million for IaaS and PaaS could lead you to think that the company is not very effective in those areas or that its installed customers are not eager to move to the cloud. That’s a logical deduction that just happens to, if not be wrong in detail, at least omits vital information.
At the recently concluded analyst event in Redwood Shores, Oracle President of Product Development, Thomas Kurian, provided insights that, while not intended to explain the revenue numbers, nevertheless shed light on the direction.
The company is planning for 13 distinct regions with triple redundant data centers around the world linked by a high-speed backbone. Each regional data center is architected to be modular and highly scalable enabling the company to increase its footprint as demand rises. Right now, of the 13 regions, only 3 are up and running. These are US West, US East, and Germany. The UK will come online soon with others to follow in the coming several years. So with 10 more regions to be implemented, Oracle finds itself in a typical dilemma. It can’t sell what it doesn’t yet have. So today, there are large swaths of the planet where demand for Oracle infrastructure can’t be met.
When viewed from this perspective, $396 million starts to look different. Three out of 13 deployments is 23 percent of planned availability. If you make the assumption that all regions could or would generate roughly equal revenue contributions to IaaS and PaaS you get a revenue number in the $1.4 billion range. That would warm the heartstrings of many financial analysts and reduce anxieties about Oracle’s uneven numbers.
Assumptions like this can lead us down a rabbit hole but this one does a good job of explaining the current phenomena. Why we might ask, have more Oracle customers not begun transitioning to the cloud? Infrastructure availability could certainly explain that. But so could the reality that it takes time to move apps, to build new, run parallel systems, and finally flip the cloud switch. One Oracle customer, Amud Krishnamurthi, who is leading a transformation to the public cloud at The Gap, in San Francisco, told me his company’s transformation process will likely take between 18 and 24 months.
On the other hand, the fact that Oracle consistently draws attention to competition from Amazon implies that it knows exactly where threats to its position lay, but equally where opportunity exists. Check Amazon Web Services year over year revenue growth.
Oracle’s transition to the cloud has already made headlines but reality can be slow to follow.
The company is currently engaged in a build-out that’s best understood as a technology diffusion process that will take time. In the US they can deploy on new infrastructure and they also have private cloud solutions that can run in-house. But much of the rest of the world is still waiting for IaaS. Both Larry Ellison and Mark Hurd have said that the process of converting the customer base could easily take a decade and the need for infrastructure highlights one important reason. It’s a marathon and they’ve barely broken a sweat.
Image credit - Oracle
Disclosure - Oracle is a premier partner at time of writing