Here at ParkMyCloud, we’re always trying to solve a problem many companies are facing: overspending on cloud. As we now support Microsoft Azure along with Amazon Web Services (AWS), we have spent significant time over the past several weeks talking with analysts, writers, and industry pundits about this problem and our vision for our customers and the ParkMyCloud platform. What is interesting in these discussions is the current size of the public cloud IaaS market ($23B in 2016) and the growth rate (28% in 2016).
The Size of the Massive Public Cloud “Utility”
Most analyst reports clearly state that AWS is the public cloud market leader, with at least 3X the size of Azure in terms of market share, generally measured by revenue. AWS has roughly 31-33% of the public cloud market, while Azure has roughly 11%, followed by IBM Bluemix and Google Compute Engine. All are growing rapidly. In fact, according to analyst reports, the “Big 4” are growing anywhere from 50-100% year-over-year. In fact, each of the Big 4 providers are growing much faster than the industry average – it is a truly disruptive, transformational time in IT.
What comes with this growth is greater adoption. According to Gartner, at the end of 2016 only 14% of enterprise workloads were in the public cloud, but by 2020 they expect that to grow to 41%. That basically means the public cloud needs to triple in size to support this workload migration to public cloud (assuming it’s not already “overbuilt” and new technologies don’t drastically shrink the infrastructure needed to support applications). At these current growth rates, the public cloud infrastructure market will be a $60B+ industry by 2020. This will be one BIG $%# market, right?
Public Cloud: The “Utility” for Enterprises
One way to think about the public cloud is as a “utility” – you can buy services on demand, just like electricity, or water, or heating. Each of these utilities are consumable – as you grow you can consume more, as you shrink you can consume less – it’s elastic. Seems like an interesting parallel to compute, databases, and storage. In the case of the public cloud, you are consuming IT-related infrastructure and services to build, test, and run enterprise and consumer applications which we consume either as an enterprise (e.g., Salesforce) or a consumer (e.g., Netflix) .
But like any utility, there is waste – lights are left on, faucets leak or are left running, and the heat running when you are not home, I am guessing you get the gist. This is why there are now consumer applications like Nest. Buildings and homes alike have ‘automated’ ways to turn lights and water off/on to reduce waste and save money and protect the environment. Why should the public cloud be any different?
Overspending on Cloud Expected to Grow
Enter the problem of overspending on cloud, or cloud waste – servers left running when people are not using them (at nights and weekends), oversized databases and servers not optimized for the applications they support, and storage volumes not being used or “lost” in the cloud. These are just a couple examples, there are many more.
This overspending is a huge problem in the cloud market.
Let’s break it down on the 2016 numbers:
- 66% of public cloud IaaS spend is on compute (servers, vm’s, instances in the cloud)
- Only 46% of enterprises monitor and rightsize their resources (that means 54% do nothing)
- 44% of cloud spend is used for non-production workloads (dev, test, QA, etc) – stuff that does not need to run 24×7
- Only 31% of companies monitor for unused storage volumes
|2016 / In Billions|
|Size of Public Cloud IaaS||$23B|
|Non-production Public Cloud IaaS||$10B|
So, we calculate that enterprises can save up to $6B by optimizing their public cloud spend in just 2016, and by 2020 that number grows to $17B.
Sounds like it’s time to use a “Nest for the Cloud” tool to reduce overspending on cloud and eliminate waste.