One of the major trends transforming the technology world is the meteoric growth of the large cloud platforms. Only a decade ago Amazon launched AWS to provide online services for other web sites or client-side application, and by doing so changed the way tech services would be served and consumed forever. Several other cloud platforms quickly rose either to satisfy vast internal demand (Facebook), to directly compete with AWS (Microsoft Azure), or to do both (Google Cloud Platform).
It is difficult to grasp how fast these cloud platforms are growing and how big they have become. This new transformation impacts almost any tech company out there. On the one hand, the new cloud platforms level the playing field by enabling startups to quickly build new services without having to spend a lot of money upfront, and to easily deploy these services at a scale which was never possible before. Yet on the other hand, the large cloud platforms are becoming somewhat of a black hole, slowly sucking in additional services as they provide more and more functionality while eating away many other vendors.
So how big is this relatively new phenomenon? Qualcomm expects that by 2018 not less than 40% of the servers shipped worldwide will be bought by one of the hyper-scale cloud vendors (e.g., Google, Amazon, Microsoft, Alibaba). This has tremendous impact on anyone who sells infrastructure-related components (compute, storage, networking etc.). First, it means that the number of potential customers is quickly declining and becoming more concentrated among several few large cloud platforms that can make or break a vendor. Second, these cloud platforms are extremely sophisticated and benefit from huge advantages due to their massive scale. Most cloud platforms don’t buy systems from Cisco, HP, Dell and the likes, but source directly from component manufacturers and assemble their systems in house.
Source: Qualcomm, 2015
IDC research also shows how the market share of traditional data centers is constantly declining. It is almost safe to assume that traditional data centers will only exist in the future for extremely large enterprises, niche use cases, or due to complex security or compliance needs. It will be almost impossible for enterprises to build and maintain traditional data centers that can compete with the quality, ease of use, and cost benefits of cloud platforms.
While there are several large cloud platforms, AWS is clearly the leader in this space. However, it’s too early to call winners and losers, especially given the early stage of the market. According to Structure Research, the world’s top cloud providers made $11.2 billion in revenue in 2015, but this is only the beginning. These cloud providers are forecasted to collectively reach more than $120 billion in revenue in 2020, growing the market more than 10-fold in just five years!
Source: Massive-Scale Cloud, April 30, 2016, Structure Research
The recent AWS Re:Invent conference demonstrated how big this new ecosystem has become. The conference now attracts the most influential developers, IT managers, vendors, and executive decision makers in the world. Amazon’s Reinvent has by now surpassed VMware’s VMwrold as the top IT conference in the world.
Source: VMware, Amazon
What’s even more interesting is that last year AWS also surpassed VMware, the leading virtualization service provider, in revenue. Up until February 2015, when Amazon started breaking out it’s AWS business unit financials, most analysts and industry experts assumed AWS isn’t making money. How wrong they were! AWS is the biggest growth driver for Amazon’s operating profits. In 2015, AWS generated $7.88 billion in revenue and $1.86 billion in operating income. These numbers represent year-over-year growth of 70% and 182%, respectively. Analysts now believe that if AWS was an independent company it could be valued at north of $150B enterprise value, which is almost 3x that of VMware.
Source: Amazon, VMWare, CapitalIQ
But perhaps most notable is the top cloud platforms’ pace of innovation. Unlike the previous generation of enterprise giants such as SAP, Oracle, IBM and the likes, the new cloud platforms are not only getting significantly bigger over time, but also getting significantly better. They hire the top R&D talent, spend tremendous amounts on R&D projects, constantly extend to new fields, and most impressive- often move at a pace of a startup despite their huge size. Any startup that offers services which can be offered by these cloud platform faces tremendous risk.
For example, at the latest re:Invent conference, AWS announced more than 20 new products and 30 large new features in almost any field out there. Most of these new products are akin to well-funded startups that have the entire AWS salesforce and ecosystem behind them. One such example is Amazon’s new Lightsail service which goes directly after DigitalOcean, a fast growing startup which raised $123M trying to compete on the lower end of the market which Amazon doesn’t yet own.
What this means for tech companies (both startups and large vendors alike) is that if you are selling systems or infrastructure software you will be facing tough times ahead. Your target customer base will constantly shrink as they migrate to cloud platforms. And it is almost impossible to sell to these large cloud platforms, since their bias is always to build and not to buy.
The best and maybe only way for startups to sell infrastructure technology in the future will be to provide tools and services on top of the cloud platforms. Such services should enhance these platforms and not directly compete with them. The new wave of successful infrastructure companies will have to acknowledge this huge trend, and try to ride it or risk getting extinguished.