The world’s favorite messaging app Snapchat’s parent company Snap, Inc. recently unveiled its initial public offering (IPO) prospectus, which gave us a clear view of how companies that deal with large volumes of data are deploying their cloud infrastructure.
For a big company that handles big data, infrastructure costs will typically run in the hundreds of millions to billions of dollars. At that level there’s a clear choice – build your own infrastructure and scale it up as you grow, or opt for a public cloud solution from a top cloud service provider (CSP).
The problem with the first option is obvious – high cost of ownership. But there’s also a catch when it comes to public cloud. At that scale and size, vendor lock-in is a very real fear because you really don’t know what direction your cost is going to take. It’s not entirely in your hands.
With 158 million people using Snapchat daily and over 2.5 billion snaps created every day, Snap, Inc.’s data requirements are on an explosion path, which will put considerable pressure on the company. Building out datacenters is an option, as we said, but it’s easier said than done. You need to carefully project your future volumes, build out datacenters with enough capacity, and be ready to add more if needed.
Aside from the planning, there’s a considerable outlay of capital required. More often than not, there’s also the problem of over-capacity versus under-capacity depending on how stable or erratic future growth is.
Public cloud is an ideal solution because of the scalability and flexibility, and the cost benefit is definitely there because there will be special pricing contracts in place for such a big client.
But what if the relationship between client and vendor sours over time? In such a scenario, Snap could find itself locked in to a single vendor, which makes moving out a tremendous task.
That’s what we believe was Snap’s motivation behind deciding to take the multi-vendor cloud infrastructure approach to the problem. The company has decided to spend a total of $3 billion on cloud computing infrastructure over the next five years, but it has cleverly distributed that between Google Cloud Platform ($2 billion) and Amazon Web Services ($1 billion).
“Google provides a distributed computing infrastructure platform for business operations, or what is commonly referred to as a “cloud” computing service, and we currently run the vast majority of our computing on Google Cloud.
Any transition of the cloud services currently provided by Google Cloud to another cloud provider would be difficult to implement and will cause us to incur significant time and expense.
We have committed to spend $2 billion with Google Cloud over the next five years and have built our software and computer systems to use computing, storage capabilities, bandwidth, and other services provided by Google, some of which do not have an alternative in the market.
Google has broad discretion to change and interpret its terms of service and other policies with respect to us, and those actions may be unfavorable to us. Google may also alter how we are able to process data on the Google Cloud platform. If Google makes changes or interpretations that are unfavorable to us, our business would be seriously harmed.
On January 30, 2017, we entered into the Google Cloud Platform License Agreement. Under the agreement, we were granted a license to access and use certain cloud services. The agreement has an initial term of five years and we are required to purchase at least $400.0 million of cloud services in each year of the agreement, though for each of the first four years, up to 15% of this amount may be moved to a subsequent year. If we fail to meet the minimum purchase commitment during any year, we are required to pay the difference.
In March 2016, we entered into the AWS Enterprise Agreement for the use of cloud services from Amazon Web Services, Inc., or AWS, that was amended in March 2016, and again in February 2017. Such agreement will continue indefinitely until terminated by either party. Under the February 2017 addendum to the agreement, we committed to spend $1.0 billion between January 2017 through December 2021 on AWS services ($50.0 million in 2017, $125.0 million in 2018, $200.0 million in 2019, $275.0 million in 2020, and $350.0 million in 2021). If we fail to meet the minimum purchase commitment during any year, we are required to pay the difference.”
With one leg inside GCP and one in AWS, Snap will be in a much better position to negotiate rates after the five-year period is over. Not only that, but it’s clear that Snap recognizes that there may be a fallout due to unfavorable changes in the GCP License Agreement.
Several things can be gleaned from this.
First of all, Google’s hidden cloud revenues are now making their appearance above the surface. We know that Google made approximately $896 million in quarterly cloud revenues during the final quarter of 2015, and the same source, New Hampshire-based Technology Business Research Inc., estimates Google’s total cloud revenues for the 2016 fiscal to be at about $4.1 billion.
The new deal with Snap, along with organic growth, could now push Google’s cloud revenues to over $5 billion for 2017.
But that’s still far less than 50% of what top CSPs AWS and Microsoft are posting in annual cloud computing revenues. What Google needs right now is a bunch of big-ticket cloud customers with long-term contracts to push its cloud revenues past the $10 billion mark. We believe that Google may not report standalone cloud revenues until then.
The second thing to note is the emergence of the multi-vendor model for cloud computing deployment. Snap is not the first company to take this route. Apple did it before, moving some of its data out of AWS and granting GCP a hefty $600 million contract for cloud services. It’s not as significant as Snap’s deal, but it does signify a shift towards preferring multiple public clouds over a single vendor.
As Snap goes into IPO and beyond, their need for cloud computing services will only increase. Snap has decided to go the Apple way and play the cloud providers off each other.
Both Google and Amazon have the money muscle to keep lowering their prices, but Amazon has a distinct advantage over the Alphabet-owned company in that it is highly profitable despite a slew of price cuts to its core compute and storage solutions.
We don’t yet know whether or not Google’s cloud business is profitable, since much of its infrastructure is used by the company’s own hungry assets, but we do know that Google will go to any lengths to prove its ability as a large-scale cloud infrastructure provider.
If nothing else, Snap’s IPO has revealed some very insightful details about how current-day companies that are growing fast will eventually deal with their infrastructure needs.
Thanks for reading our work! We invite you to check out our Essentials of Cloud Computing page, which covers the basics of cloud computing, its components, various deployment models, historical, current and forecast data for the cloud computing industry, and even a glossary of cloud computing terms.