The real winner behind Snap’s IPO

Amit Karp
Venturing startup nation
3 min readFeb 8, 2017

The big tech story of the week was the highly anticipated Snap S-1 filing release. But besides revealing user growth, revenue, and (partial) engagement metrics, the big takeaway from the filing was Snap’s deal reached with Google a few days ago. As part of this partnership Snap committed to purchasing $400 million in Google’s cloud services annually until 2022, totaling $2 billion over the period.

Cloud-based infrastructure services started only a decade ago and have grown at meteoric pace. However, Snap’s S-1 filing so clearly revealed the magnitude of this phenomenon. We are witnessing a relatively new fee, or a “cloud operating tax”, which every technology company (or, in effect, every company) will end up paying the cloud platforms.

Almost any technology company depends on cloud platforms for at least some of its infrastructure and services. This dependency keeps growing over time, trending towards completely relying on the large cloud platforms for anything infrastructure related. It is almost safe to assume that traditional datacenters will only exist in future for extremely large enterprises, niche use cases, or for complex security or compliance needs. Even the most advanced technology companies are slowly caving in and moving more of their workload to public cloud platforms. Only last year Uber said that it will outsource infrastructure to the public cloud for the first time after seven years of existence, Spotify announced that after years of operating its own datacenters it is migrating the whole infrastructure to Google Cloud Platform, Netflix revealed that it is not running any more datacenters after finishing a seven-year migration to the cloud, and even Apple signed a deal for between $400 and $600 million with Google Cloud Platform.

The reasons why all these tech players are abandoning their own infrastructure and migrating to the cloud are always the same: today it is almost impossible for anyone to build and maintain traditional datacenters that can compete with the quality, ease of use, and cost benefits of the large cloud platforms. This becomes even more significant over time as these cloud platforms keep advancing at a staggering pace. We will therefore likely be left with mainly three large cloud vendors available in the US: Amazon Web Services, Google Cloud Platform, and Microsoft Azure (unless the Chinese players enter the US market).

What this means is that every tech company or any company that delivers digital services will end up paying one of these three cloud providers an ongoing annual fee for using their infrastructure and services. This fee (or “cloud operating tax”) will be paid for every digital service consumed around the globe. I don’t believe we have ever seen anything like this before. How high is this tax? It really depends. Snap’s cloud spend compared to its current revenue is staggering, but it will decrease over time as revenue increases. However, it is not uncommon for several of our companies to spend 20% of their revenue, if not more, on cloud platforms.

The main three cloud ‘tax’ collectors benefit from two trends: the ongoing growth of the amount of data, traffic, and technology services consumed over time; and the constant migration of existing workloads from traditional datacenters to the cloud. It is therefore not a surprise that these three companies are among the top five valued companies in the world. Collecting an ongoing ‘tax’ from any technology service delivered over the world for the foreseeable future while having limited competition is an incredible business model to have.

--

--