The Tipping Point for IaaS
The Tipping Point for IaaS
By Rob Bissett, 6fusion SVP of Product Strategy
There has been plenty written about the flurry of cloud pricing moves started by the Google Compute Engine price drop last week. Most of this comes down to analysis of the facts themselves, and the impacts on current competitors and the market as a whole. However, when you sift out the signal from the noise, a much bigger story emerges than just the cascade of a few price changes.
What are we talking about? We are likely on the verge of something the industry has been talking about for a long time, but that seems to be coming much faster than most people anticipated. I am calling it now – I believe what you saw last week was the tipping point for the commoditization of web scale IaaS. The rationale for my thinking is outlined below, but the TL;DR version is this: the IaaS market hit an inflection point last week and is forever changed.
First – why do I think we have started “the next phase” of this market development?
Reason #1 – We now actually have a commodity. With the movement of GCE to general availability and continuing moves by Joyent, HP Cloud, and now Microsoft to beat the drum for web scale IaaS services, we finally have a category that can meet the true definition of a commodity – a good or service with no qualitative differentiation. We now have a critical mass of major, brand-name recognizable, stable (er, well mostly), and global firms that are all touting infrastructure services that look, smell, taste, and feel about the same.
Reason #2 – With continuing advancements in APIs, and the use of deployment and management tools like, Chef, Puppet, Rightscale, ServiceMesh (and about a million others), a firm or a developer can use one or more IaaS suppliers, in one or more regions, and actually not notice or care who the supplier is. That’s big.
Reason #3 – The flurry of recent announcements of new value added services provided by the big IaaS vendors has really shifted the argument about where the value lies from these large IaaS services. Now that virtually all of these companies are providing DNS, load balancing, database, and other services on top of the IaaS stack, the competitive value add has moved to ever better and more advanced services. The large IaaS providers will continue to innovate on the back end, but in the future will drive their external competitive differentiation on the value added services. Using the underlying IaaS services to drive consistent margin and to act as a gateway drug for the more profitable and differentiated value added services – exactly as happens in virtually all commodity markets.
Reason #4 – For a long time Rackspace has doggedly chased Amazon as the #2 vendor. With the entrance of Google, and the continuing growth of Microsoft and HP in this market, the market power dynamic has started to shift. Though market share numbers are still strongly in Amazon’s favor, we finally have a viable race for a #2 that can exert market pressure on #1. The proof? The recent Google price cut. With that move, Google was able to do something that no one else had yet accomplished – force a counter move by Amazon. While Amazon has a long history of driving pricing down, to date they have done it “largely without any competitive pressure” to quote Andy Jassy at a recent AWS Summit. To me, this represents a significant market signal that Google is for real and Amazon is taking them much more seriously than other competitors. Remember, Google doesn’t have to have as much market share as Amazon to impact the market, they just have to have enough to be a threat. Google, HP, and Microsoft collectively now appear to have that power and we will see market dynamics start to accelerate as a result.
Any of these factors taken on their own are really just a signal of a maturing market. Taken together, in such short order, signals something much more important – that the emergence of the IaaS commodity is near.
Look – commodity isn’t a bad word. Virtually all of the products and services that make our economy run can be classed a commodity. Gasoline, electricity, fruit, meat, and many, many others. These types of commodities make it very simple for buyers to determine how much they need, how much it costs, and lowers (dramatically) the friction involved in purchasing the good or service.
To support this, for each type of commodity a framework (or market) is developed with standard units of measure, standardized contract terms and the like. Why? Well its simple – it makes purchasing these services and goods simple and cheap. Retail markets for these goods still exist, and there are many specialist services that surround and support them to help users consume them, and to enable users with specific requirements or tastes to purchase particular services, and these retail services never go away.
Why is this good? Pricing. By creating standardized services with multiple providers all able to deliver these services, the cost and risk of acquiring services for users drops dramatically, resulting in a massive increase in the adoption of these services.
For suppliers, the cost of service delivery and business acquisition will drop dramatically, as scale continues to grow. Suppliers will create enormous economies of scale, drive internal costs as low as possible, and innovate on value add services to drive increased profit margins (sound familiar?). All of this is the end result of the maturation process for the market, and only happens with those services that are critical to our economy, and I would suggest that cloud clearly meets these requirements.
So what happens next? Clearly we aren’t there yet, but I would argue that we are well on our way, and that the progression is both valuable to the market as a whole, as well as inevitable. That being said, there are some things we (the market) need to do to get there. I believe that the race to provide abstraction and management will continue, and is critical to reducing the technical risk and friction involved in cloud adoption. The continuing work on fungibility, while helpful, won’t decide the future of this market. Finally, the biggest thing we will need to develop is the actual market around IaaS services. To date, we have seen the emergence of many “brokers”, “retail marketplaces”, and the like. I will refer to all of these as “disorganized markets” in that they are ad-hoc, not standards based, and operate without a regulatory or industry standards. We will see a huge shift in how this works over the short term, with the emergence of organized markets marking the next step in the evolution of the IaaS market. How will this happen? Stay tuned for the next episode…