As Google and Amazon race to the price leadership bottom in the grab-customers-at-all-cost push, one wonders if a “Cloud Bubble” isn’t forming. Over the past year, both Amazon and Google have announced price drops in excess of 50% for select services in an attempt to capture cloud share. Microsoft, not be outdone, has announced a push towards Office 365 giving out substantial discounts to attract customers. All three have increased the amount of customers using cloud substantially. Research by Synergy Research Group of the big 4 cloud providers shows Amazon holding the greatest market share owning 27% of the cloud market. Microsoft is second capturing approximately 10% with IBM owning 7% and Google grabbing 5%. But with many analyst speculating that the margins in cloud are at best in the single digits, and potentially in the red, one has to wonder if a bubble is forming. How inexpensive can the cloud become and is this movement setting up a potential burst? To understand what the bubble will look like, or more importantly, what will be the impact when it burst, let’s take a look at a potential scenario that could occur.
CIO’s recognize the need to streamline and automate their data center to take advantage of the efficiency, agility and unification that only a private cloud provides. It’s not enough to stop at only virtualization in the data center. Many enterprise data centers are examining and developing private cloud infrastructure to ensure the integration of their legacy applications to their cloud enabled and SaaS applications. Once private cloud adoption moves into wide stream adoptions, public cloud growth will slow and public cloud providers will need to determine a way to turn their existing customers into long term profitability.
Now, in order to create long term profitability, they’ll need to increase prices. And this is where it will get complicated. As public cloud costs rise, enterprise customers will look to move applications back into their private cloud. The ease of this migration will hinge on exit cost analysis; how simple is it to move applications back into the private cloud data center. And exit costs will directly correlate to the degree of integration with the public cloud provider’s system. Customers running applications on a Microsoft, VMware or OpenStack cloud stack will potentially have a lower exit cost, provided they are running the same release version. Customers who have designed their applications to run exclusively on EC2 or who are taking advantage of Cloud Apps engine may be looking at rewriting applications to run in private cloud environment, and unplanned and unbudgeted exercise. Intuitively, Amazon and Google have the least to lose by increasing cloud pricing. They have entrenched customers who have high barriers for exiting their platforms.
Does all this sound vaguely familiar? One only needs to take a look at the mainframe era to recognize the similarity to applications directly tied to compute environment for which they live to see the parallel. So, how does all this shake out? Smart CIO’s will begin to recognize the potential ramifications of too tightly coupling their applications with their cloud providers. And it is this insight that will lead to many enterprises ensuring the movement of their applications to a private cloud. As public cloud providers increase cost, barriers to exit will come down accelerating the exit. At this point, some public cloud providers will not be able to support the diminishing margins. This is the point when cloud companies who cannot financially weather a declining public cloud use, will implode. And the bubble will burst.
We have seen the effect of what happens when cloud providers close shop. Nirvanix and MegaCloud are just a couple of examples reference in the NetworkWorld article posted June 2nd 2014. So, what shape will you company be in if the bubble burst?