It takes time to make a cloud
Financial analysts are in a small funk because Oracle’s earnings were not as spectacular as they’d have liked. Actually, the revenue numbers were fine; they beat the street estimates of $9.57 billion for the just completed Q2 2018 with revenues of $9.63 billion. The cause of the consternation was a relatively weak cloud growth rate of “only” 44 percent after posting 51 percent growth in the prior quarter. That’s a downward trend and thus the consternation. What explains this?
Let’s focus on human nature. Financial analysts look for eye-pleasing graphics and numbers that tell wonderful stories of increase. But the reality is always more complicated. Orders don’t ship, customer CTOs get cold feet, CEOs find new bright and shiny objects to pursue. Stuff happens. As a result, often a vendor’s numbers don’t’ look as impressive as we’d like.
One specific area of concern for Oracle has been the speed at which the market is adopting its IaaS and PaaS (infrastructure and platform) product lines. The SaaS line seems to be good. Oracle sold $1.1 billion of SaaS in Q2 making it one of the biggest SaaS companies on the planet. But it’s still struggling with infrastructure and platform which combined brought in “only” $396 million in the quarter.
The SaaS business seems to be new deals won the old fashion way. But it takes more effort to grow the other two because much of the increase is logically expected to come from Oracle’s customer base. Analysts are expecting existing customers to swarm in to the new offerings but they seem to be taking their time. The maxim, if it ain’t broke, don’t fix it comes to mind.
Most enterprises have a huge investment in hardware and software, which they are naturally reluctant to discard. So while Oracle’s offer to let them migrate their existing licenses to the cloud in a program called BYOL or bring your own license, that offer is not sweet enough for many businesses with gear they’re still writing off.
At this rate it will take more than the efficiencies from better infrastructure to motivate many businesses to move, hence the disappointing numbers. But to be a bit more realistic, customers are moving to the cloud—$396 million is far from nothing. But it’s going to take more time than the optimists thought to get the migration moving at a faster clip.
At this point there are many things Oracle can do. First, it should count and publish the numbers of businesses coming to their infrastructure. This will likely represent customers gaining a toehold in the cloud. They’ll bring one application or department to the cloud as an experiment and Oracle should pay attention to this because it’s one sure method of growth.
Even if the revenue numbers are small the absolute number of businesses going to the cloud, even if only partially, will strongly suggest future acceleration. For all we know such information is already embedded in the $396 million Oracle already reported. Second, Oracle can sell platform. They do this already but perhaps some effort more squarely aimed at the small developer groups that just need a sandbox. That’s where early growth often comes from. Those groups are often oriented to Microsoft products running on AWS so the strategy works well in two directions. Lastly, when all else fails, the financial analysts could always try being a bit more patient.
My 2 bits
Moving to the cloud is an arduous event for any company. It calls for managers to tinker with the secret sauce, something they loathe. So it’s no surprise to me that moving is a slow process and that despite having all of the bells and whistles ready to go, Oracle is still in early market hurry up and wait mode. It can’t be helped—no vendor can expect to sell anything until it produces the whole product. Oracle CEO Mark Hurd is famous for saying they’ve built a skyscraper but couldn’t start renting units until the building was finished. That’s the time when financial backers begin asking about revenues.
But it takes more time than we care to admit to make a cloud. Oracle is showing some promising signs of early adoption but the situation still calls for patience.
Patience? What’s that?
IaaS Market—Who Wants It?
I saw this headline on Yahoo Finance, “Oracle Has Big Plans To Beat Salesforce And Amazon In A $72 Billion Market” and it got me thinking about a lot of things.
First, the market in question is Infrastructure as a Service or IaaS and I didn’t think that Salesforce was actually in the IaaS market. This shows how far we still have to go in cloud computing just to get the terms right and set the playing field.
IaaS is all about computing infrastructure — the servers, memory, disks, operating systems, middleware, databases and maybe applications — that any modern business needs to support its information processing when it chooses not to manage all this by itself. Companies subscribe to IaaS services and it’s like Walmart for conventional IT. You know this.
Second, IaaS is part of a cloud industry that also includes software as a service or SaaS — systems that are centrally managed and subscribed to and PaaS or platform as a service. PaaS is the top of the heap right now, the place you go when you don’t want to manage the gear as in IaaS but still want a robust place to design, develop and manage applications that are embedded into your business processes AND to generate those apps for multiple platforms from a single design.
Interestingly, a PaaS provider probably, but not necessarily, is an IaaS provider by the simple logic that the “P” is software and it has to live somewhere. So this brings us to the headline and article and my issues with it.
Oracle has some very nice server, storage and in-memory gear that it acquired when it bought Sun or built once it had the Sun gurus on board and working with its database mavens. It can very well supply the IaaS market — those organizations that want to run their own IT shows but not operate their own data centers. Great! Amazon is in that business. Google is in that 0business. Oracle is in that business and it is trying to be in the PaaS business with its still-being-delivered Fusion platform. Salesforce is in the PaaS business but not really in the IaaS business except by accident of circumstance noted above.
The article quotes some succulent swipes between Marc Benioff, CEO of Salesforce, and Larry Ellison, CEO of Oracle, and Marc’s former boss. The quotes are old and largely irrelevant but you have to admit, these guys sure know how to grab a headline.
The terminology wars
I had the pleasure of attending two events last week, Microsoft Convergence 2010 in Atlanta and the Salesforce-VMWare VMForce.com announcement in San Francisco. Each event was useful and informative in its own way and I am pleased to have witnessed each first hand. I have written exhaustively — at least I’m tired — about each of these so rather than another recitation of events, I want to spend this piece drilling into one aspect.
Terminology was common and different in the two events and it’s worth looking at. The term de jour is cloud computing and it has been current for at least ten minutes — more like about two years. We know it’s the “in” thing because until very recently both Salesforce and Microsoft were using it albeit in very different contexts.
A couple of years ago, Salesforce and a few other intrepid vendors began using cloud computing to describe the confluence of SaaS applications, increasingly available infrastructure and the ability to “mashup” components to make novel solutions. Cloud computing meant Infrastructure as a Service (IaaS), SaaS or software as a service and PaaS or platform as a service. For instance, Google Maps, running on Google’s server farm, plus your prospect list in SFA, running on Salesforce’s farm, could yield a map or satellite photo of a territory with markers indicating prospect locations. In one vivid and visual screen a person could see and strategize a set of sales calls, for instance.
Cloud computing was fun and it replaced SaaS, more or less. To be SaaS seemed to imply an application ran for the benefit of those who were earthbound in a building or within the constructs of the enterprise. To be in the cloud was to let your imagination soar, to combine and recombine in limitless ways to produce new functionality, to go where no man has…er, you get the idea.
Cloud computing was fine and wonderful until many vendors either entered the space with their own ideas or when they discovered that they couldn’t exactly do it without serious reconfiguration of their product sets and business models. Cloud computing was a disruptive innovation.
Not to worry, the inevitable commoditization of the term then began in earnest. Like watching the sky on a windy day clouds could morph into wonderful shapes and thus cloud computing could be many things. That’s about when Microsoft entered the fray with Steve Ballmer’s famous quip that “we’re all in” meaning that the software giant was making the biggest bet that it could on cloud computing.
“All in” for Microsoft means that their vision of cloud computing is essentially infrastructure as a service (IaaS) — the ability to tap into a server in the sky on which you’ve bought space to run a conventional application. With Microsoft’s market power it is likely that cloud computing will come to mean whatever the company wants it to mean and that means cloud computing will look more like time-sharing than a mashup.
This brings us to last week. While Microsoft was all in, in Atlanta, Salesforce was way out on the west coast announcing an alliance with VMware. You’ve seen the news no doubt, that the Force.com platform is being enhanced to provide Java runtime services for legacy applications through this association. It demos well and if it runs that way I think we’ve got another important brick in the wall of new-fangled IT.
In his introduction, Salesforce CEO, Marc Benioff, called the alliance the beginning of a new era in “Mobile Internet Computing”. When he uttered those words my ears pricked up and it made me wonder if we were witnessing the birth of new terminology yet again.
Salesforce and its close allies and partners have managed to stay one step ahead of the terminology commoditization curve for the last ten years and it wouldn’t surprise me if we were about to get another new term. In this way, Salesforce has managed to use accelerating terminology commoditization to its benefit saving its marketing budget for more important things and always managing to pigeon hole its competition as retrograde.
I have only heard Mobile Internet Computing once and it may not stick but I wouldn’t be surprised if it did. The thing I don’t understand though, is why Microsoft continues to fall into the same trap of letting someone else name a trend and then spending lots of its own money popularizing it.
An alternative for Microsoft might go like this. Microsoft has four ERP systems and one CRM system and you could make a good argument that it is an ERP company and that ERP ought to be an on-premise application except in rare cases. I am not saying I agree with it because there are other benefits of cloud computing we’re not talking about here.
But the ERP worker is usually someone who works in the finance or a related department. This worker does the job at a desk in an office and has little use for a SaaS application. So why the big push to call this cloud computing? Selling infrastructure services isn’t really cloud computing so why call it that? Why not define yourself in terms of what you do rather than in terms of what others make up?
If Microsoft wants to compete in cloud computing CRM that’s fine. It could easily develop a hybrid model for integrating front and back office applications and not have to fight an uphill battle with its ERP customers over where their data is stored. It would be a novel position and one that is immediately appreciated by its ERP customers.
