“Price cuts that instigate price cuts” is probably a good way to describe the current cloud wars with vendors bringing a whole new meaning to the term “flexible” and bending over backward to be more accommodating. Offerings that are just so lucrative that competitors are seemingly compelled to respond with one of their own is what we’re talking about here. “Increased flexibility” seems to be what everyone is after and every new cloud pricing option from a major cloud vendor is usually accompanied with the announcement that it’s more flexible than the competition.
Price is outdated
Google Cloud is catching up quickly, and recent studies cite GCP as the slightly cheaper option, but the king of the hill (and the king of price cuts) is definitely AWS. By July of last year, AWS had its 62nd price reduction, which says a lot about a technology that is still relatively new. Microsoft isn’t far behind and cut prices of some of its offering by up to 51 percent earlier in the year.
Just as everything in the enterprise is changing, from the way we develop to the way we deploy and even to the way we monitor, the cloud wars have changed, too. The cloud pricing wars aren’t even about price anymore, mostly because price is just an outdated term now. There used to be a time you could look at the providers’ offerings side by side and compare prices, but that’s pretty hard to do now, especially since most offerings look like a line of algebra followed by a price tag.
Flexibility wars trump cloud pricing wars
Looking at cloud offerings today, the options for configuring your plans vary so much that it’s not the prices that matter now, but rather to what extent a vendor lets you custom tailor your usage. That’s what all the talk about flexibility is about and the more you are able to customize and configure your usage, the less likely you are to end up paying for stuff you don’t need or aren’t using.
The current war is all about letting you use the cloud any way you want. Google recently announced new Committed Use Discounts (CUDs), which give customers a reduced rate on virtual machines in exchange for a long-term commitment. While Azure has enterprise agreements and AWS has reserved instances, the difference here is that users don’t have to commit to what kind of virtual machine they will be renting.
With CUD, all you need to do is estimate your total usage over the life of the contract. AWS was quick to respond as usual and announced Instance Size Flexibility for Reserved Instances. This was obviously a response to Google trying to sell the fact that Reserved Instances aren’t “flexible” enough, and AWS customers are now able to change the virtual machine type they’ve reserved. To put an exclamation mark on the reply, AWS is also offering Convertible Reserved Instances with even greater flexibility at a slightly higher price.
Microsoft Azure recently announced Reserved VM Instances, which means users can now purchase advanced capacity and save up to 72 percent compared to the on-demand price. Additionally, users that incorporate Azure Hybrid Use Benefit to transfer Windows Server licenses to Azure get even bigger discounts. While letting you customize your usage is a big part of the current cloud pricing wars, billing is the other big part.
Billing by the second
While AWS was billing its customers by the hour, both GCP and Azure introduced billing by the minute. Starting October 2, AWS customers with virtual servers running Linux were billed per-second, with a one-minute minimum usage. Amazon’s Jeff Barr in a blog post said, “Many of our customers are dreaming up applications for EC2 that can make good use of a large number of instances for shorter amounts of time, sometimes just a few minutes.” The move is also aimed at Lambda, where instances are triggered by events, sometimes for just a few seconds.
Google kind of stole the air out of that announcement — and in more ways than one. In just a matter of days after AWS’s declaration, Google announced that its cloud service will offer per-second billing and made it a point to state that its customers will feel less impact from the change than users of a certain unnamed vendor that used to charge on a per-hour basis (we wonder who). Additionally, Google is also filling in the gaps AWS left in its game by offering customers per-second billing regardless of whether they’re using Windows or Linux. Additionally, the new per-second pricing will apply to most of Google Cloud’s core products, including Compute Engine and App Engine.
Oracle can be flexible, too
Oracle may be a distant No. 6 behind market leader AWS, but that hasn’t stopped it from vowing to knock AWS off the top spot. Mark Hurd, Oracle’s CEO, said in an interview with CNBC at the OpenWorld conference in October, “I like our chances against anybody — anybody.” To hammer home the point, he added, “I think we’re going to win, bar none.” Oracle joins the “flexibility wars” by releasing a much more flexible consumption model that will both cut costs and enhance performance. This new model is being called Universal Credits, which comes with the added advantage that it contractually ensures customers have access to all future PaaS and IaaS features.
With Universal Credits, the customer gets one single passcode or SKU that gives access to the entire suite of Oracle’s PaaS or IaaS and lets them activate a new service any time they want. Oracle has also introduced BYOL (bring your own license) to PaaS, which enables customers to use their existing licenses (like on-prem) to access the equivalent PaaS on Oracle cloud for a fraction of the cost. Additionally, customers can change or cancel services at will without notifying Oracle and apply those credits to new services, pretty much like a self-service cloud.
After the announcement, Oracle executive chairman and CTO Larry Ellison explained that by letting customers help themselves, the human element in the PaaS process can be eliminated and the savings would be enough to halve Amazon’s cloud costs. Oracle also boasts that running Oracle Database on Oracle IaaS beats Amazon on speed as well as a variety of features.
The bottom line
Competing in the cloud is hard, and while the race has become less about direct price cuts and more about pricing models that give users a variety of ways to configure workloads, the bottom line is still money. The whole business is based on making unlimited supercomputing power available at unbelievable prices so that all kinds of users can be accommodated. This means big investments and ramping up datacenters so that it’s actually possible to rent out virtual computers for pennies by the hour.
As the cloud wars change from a rush to the bottom in terms of cloud pricing to a rush to being the most flexible in terms of usage and billing, distinguishing your offering from the competition will become increasingly difficult. In the end, it will be the easiest vendor to deal and negotiate with who wins this battle, and the users who will get a much better overall experience as a result of it. So, in the cloud pricing wars, it will ultimately be the consumer who emerges victorious.
Photo credit: Freerange Stock