- Google’s efforts in the cloud aren’t very well understood.
- However, when breaking down the pricing/feature set, it seems Google has identified a niche to cater to.
- Amazon is still the best bet, though the other two are worth keeping an eye on.
The triggers for driving cloud adoption are security, pricing, features and support. The Alphabet Inc-C (NASDAQ:GOOG) Google Cloud Platform (GCP) is considered the cheapest of the trio but lacks in features and support. Google’s low price point is driven by per-minute billing, a sustained use discount (the pricing declines the longer you run a node in Google’s data center), and pre-emptive virtual machines (VMs) that offer temporary workload instances (around 2 to 3 times cheaper than full-service VMs), but uses spare capacity (and can be turned off at a moment’s notice) which implies that a server instance that needs 100% uptime will not be a good fit with Google’s pre-emptive VM.
Google’s network of applications that run on top of its service is rather small when compared to both Amazon (NASDAQ:AMZN) and Microsoft (NASDAQ:MSFT). In other words, many of the enterprise apps that are offered in Amazon's AWS aren’t available on GCP. AWS has 400+ services, which compared to 20 on GCP. However, Google makes up for this with a Platform as a service (PaaS) approach that caters more specifically to app/web developers as opposed to business executives by creating a more streamlined process for setting up a workload instance and providing tools for application developers to jump in and develop apps.
Here’s how Deutsche Bank describes this process:
Developers can deploy applications quickly by having Google host various developer tools and set up the server and infrastructure environment. Developers simply create the code and describe to Google how they would like the application to run (specifying run-time requirements, the type of OS, etc). App Engine then configures the VMs, installs the database servers, and everything required to put the application into production.
As one customer described, the Google App Engine PaaS has “a higher level of abstraction” compared to Compute Engine, meaning that App Engine automates more functions for developers and creates more of an “out-of-the-box” experience. This can reduce the number of internal engineers required compared to either Compute Engine or AWS EC2. As an application hosting platform, customer usage of App Engine tends not to scale up/down at the magnitude that usage of Compute Engine can.
GCP is a $400 million/year business whereas AWS and Azure are $9.62 billion/year and $1.544 billion/year businesses according to Deutsche Bank. This is based on their current and most reliable estimates. GCP is quite small when compared to Amazon and Microsoft respectively. Furthermore, RBC Capital Markets estimates that Microsoft Azure generated a gross loss of $687 million in FY'15. It's hard to determine whether GCP contributed positively to GOOG's net operating income, but given the lack of scale/size and front-end development cost for various data center technologies, the segment is likely losing more money than it's generating in revenue. On the other hand, AWS had a segment operating income of $1.863 billion, and an operating margin of 23% in FY'15. For an infrastructure play, Amazon Web Services shows the most promise.
Business mix is another concern. Amazon Web Services has the most potential to swing margins and revenue in the foreseeable three-year period. As a proportion of revenue and earnings, AWS alone adds significantly to Amazon’s current valuation. Microsoft and Google both operate large revenue/margin contributing business units, so the incremental addition of IaaS/PaaS to its business mix is a rounding error, whereas Amazon’s AWS unit provides enough operating leverage to offset the weak margin profile of its retailing operation. Therefore, the exposure to cloud via Amazon is more compelling, but the weak balance sheet and fundamentals, when compared to Microsoft and Google should also be considered as well. Not that I’m saying Amazon is mismanaging its money, but both Microsoft and Google have way more money sitting in the bank, and have very acceptable asset/liability ratios when compared to the vast majority of S&P 500 companies.
The quality of GCP's support and sales lags behind AWS and Azure. Implementation requires much more technical expertise as back-end support isn’t available until you reach the higher pricing tiers, which negates the pricing discount Google provides. Furthermore, the company has yet to develop a partnership ecosystem to leverage some of GCP’s unique features. The lack of trained/coordinated sales staff to smooth the transition for enterprise limits adoption from mid-market/large firms.
Google caters to start-ups as opposed to large enterprises, as smaller firms are more sensitive to pricing and are able to live without technical support. In other words, Google is a great introductory product, but when it comes to commercialized products many prefer AWS due to the well-established history of workload compatibility, server up-time, speed and elasticity of hyper scale compute resources and a well-established history of price reductions. Furthermore, Amazon continues to innovate in the space and provides new feature improvements to speed the deployment of applications, reduced cost of ownership and offers greater compatibility with already existing workloads.
There’s still plenty of room for Amazon, Microsoft and Google to co-exist alongside each other. Furthermore, it has been noted that Google’s cloud product is still better than what’s offered by IBM, according to industry experts. So, competing with old IT hasn’t been Google’s greatest problem. Rather Google faces challenges with reaching economic break-even at price points below Amazon.
The level of internal executive support for GCP is small relative to the Search, Android, YouTube business units. It’s worth keeping an eye on GCP, because the business could swing assumptions on growth in future years. But, for the time being, investors can happily focus on Google’s efforts in its DoubleClick Ad Network, Android and Google’s consumer-facing web applications. These units could see inflection to much higher growth rates assuming a strategic shift in monetization or development of new technologies.
Amazon's positioning in infrastructure cloud and eCommerce is fairly compelling as growth in operating/audience metrics supports upwards bias to sales/earnings. The valuation is a little stretched, but has come off a little bit following the prior earnings report, so I continue to rate Amazon as a buy.