Amazon Web Services blinks and the market quivers. Microsoft shrugs and heads turn.
As the ecosystem has grown, two cloud companies control the tenor of the infrastructure as a service market. In the U.S., add in Google's push to increase market share and the cloud has the illusion of choice.
A cadre of consulting firms vie to fill the implementation gap.
Third-party solutions providers attend to the channel.
And companies implementing cloud services know only to ask for one of the big three.
The $50 billion market prioritizes the hyperscalers at the top, neglecting companies in the shadow of giants who offer alternative cloud solutions.
Analysts say the long-term value proposition of niche firms subsisting on individual developers and SMBs is lacking. The alternative providers say otherwise.
The most recent data from Gartner in the public cloud IaaS includes market share and revenue from 2018. Amazon accounted for 48% of the $32 billion market; Microsoft held 16% and China-based Alibaba, 8%.
IaaS public cloud market share, 2017-2018
|Market share, 2017
|Market share, 2018
Morsels of cloud revenue are left to alternative providers, including 23% for the "others" category — the companies that failed to make a blip on the revenue market share. Each holds less than IBM, which owns 1.8% of the market, according to Gartner.
There is not a lot of room for other providers to make a meaningful dent in terms of market share, Raj Bala, senior director analyst at Gartner, told CIO Dive. "Ultimately this is really a three-horse race in much of the western world."
Providers without the resources of Oracle have less room to grow in the market. The IaaS space requires large-scale hardware investments — providers have to grow infrastructure to meet demand and keep down latency.
"Ultimately this is really a three-horse race in much of the western world."
Sr. director analyst at Gartner
Small vendors use colocation providers as a workaround for building proprietary data centers, leaning on the physical infrastructure of Equinix, Cyxtera, Iron Mountain or others.
"The future is bleak, essentially, for those providers who do one or two specific things," said Bala. Some of the alternative providers that have emerged have not done well selling to large enterprises. DigitalOcean and Vultr focus on software developer customers deploying consumer-type applications.
Ultimately there's not enough money to be made by selling to those types of providers to create a meaningful business, he said.
Alternative providers want to carve out a niche, capitalizing on a mere morsel of the IaaS market. Gartner estimates IaaS revenue will reach $50 billion in 2020. A company owning 1% of the market would have cloud revenue of $500 million.
Any company compared against Amazon, Microsoft and Google — each of which has held a $1 trillion market capitalization — looks small, almost insignificant. But alternative companies measure success in millions, not billions. They may not win the market, but they will continue to excel.
"It's such a big market; there's so much room," Blair Lyon, VP of marketing at virtual private server provider Linode, told CIO Dive. Small providers are a "tiny part of the market but growing nicely."
It just depends on the type of customers the alternative providers are trying to serve.
Illuminating the others
Gartner doesn't disclose which vendors are included in the "others" section of the IaaS public cloud market report, but the analyst firm's Magic Quadrant on cloud IaaS offers clues. The 2019 report included Oracle, a company which has failed to have great market penetration since debuting public cloud IaaS offerings in 2015.
Oracle, which reported an operating income of $13.5 billion in FY 2019, is expanding its infrastructure footprint with the addition of five regions; its cloud runs in 21 independent locations. The company's goal is to have 36 cloud regions available by the end of 2020.
More clues about the mysterious group of others lie in the archived IaaS Magic Quadrant from 2017. The report included eight niche players: Skytap, Joyent, NTT Communications, Interoute, Fujitsu, CenturyLink, Rackspace and Virtustream. They disappeared from more recent reports as Gartner added criteria narrowing the category to hyperscalers with integrated IaaS and PaaS offerings.
Gartner found customer evaluations focused on strategic vendor adoption with broad use cases, according to the report. Some customers want "scenario-specific providers," which are best evaluated in context of a given workload rather than a larger market.
Companies forget about other clouds offering IaaS, which go head-to-head with the hyperscalers, said Melanie Posey, research vice president and general manager at 451 Research. They basically "dropped off the map in terms of competing at the same scale."
If a company doesn't need the "associated bells and whistles" of a hyperscaler, they might choose DigitalOcean or French cloud computing provider OVH, said Posey. Or they're an existing customer of IBM or CenturyLink and choose to remain with their primary technology provider.
Providers such as Rackspace and IBM have jumped on the hybrid multicloud, capitalizing on the in-between market where companies want to maintain existing systems while dabbling in the cloud.
Some alternative providers predate the cloud boom and have settled into a core customer base.
Linode launched in 2003, three years before Amazon deployed Amazon Web Services. As the cloud gained momentum, the smaller provider grew in kind. While aspects of the cloud will become simpler (industry has by now figured out storage) hybrid multicloud and pushing computing to the edge layer on complexity.
That opens up the market for alternative cloud providers, Lyon said.
To succeed, Lyon said alternative providers to hyperscalers AWS, Azure and Google Cloud have to do about 80% of all major workloads: core primitives, compute, storage, databases, bare metal, virtual private networks, global infrastructure, on-demand support … the list goes on.
The last 20% is reserved for specialized use cases hyperscalers provide, such as gaming, machine learning or quantum computing, which the average company does not need.
Alternative cloud providers aren't where Nike or Johnson & Johnson are going to put all their workloads, but they're a "great multicloud option for specific things," whether that's a testing environment or failover, Lyon said.
The other cloud value proposition
Enterprise contracts are higher value: It's no longer surprising to see large companies ink $30 million annual commitments with AWS, Bala said.
While there's a reasonable market for consumer-facing, developer-focused cloud services, there is not a $5 billion to $10 billion market opportunity if a company only sells 3 to 4 virtual machines to a software developer.
Compare any two players in the cloud market and they have inverse properties, according to Bala:
Microsoft struggles to appeal to startups, AWS excels
Google doesn't do well with enterprise, but it's Microsoft's core focus.
Though DigitalOcean and AWS focus on developers, AWS does better with enterprise providers.
Even though hyperscalers offer everything an alternative provider does, there's a niche to fill. The others want to capitalize on the independent and SMB market they say is neglected. They also pitch services as a way to avoid lock-in at a cheaper rate.
Linode has about a million customers worldwide, many of which are smaller developers and developer teams. It counts Comcast and other Fortune 50 companies as customers, but Linode is not their core infrastructure provider. It operates in 11 regions worldwide and plans to add two more — in South America and Europe — this year. Half of its customers are in North America and the rest are international, with a large portion in Europe and Southeast Asia.
Cloud infrastructure provider DigitalOcean too fills a niche. It has a half a million customers using its services as a foundation for building applications using its signature cloud Droplets.
"There is a need for other cloud providers for sure. As the big three continue to get bigger and bigger and bigger, I think there's a large slab of the of the market that is going to be underserved."
VP of engineering at DigitalOcean
Its core customer base is comprised of developers and SMBs, from hobbyists experimenting with technology to businesses after simple and predictable pricing models, Al Sene, VP of engineering at DigitalOcean, told CIO Dive.
"There is a need for other cloud providers for sure," said Sene. "As the big three continue to get bigger and bigger and bigger, I think there's a large slab of the of the market that is going to be underserved."
If a business is generating $100,000 a year, it's not going to gain the attention of large services providers, he said.
DigitalOcean recently announced a restructuring plan, resulting in the elimination of some roles, the company told CIO Dive in a statement. DigitalOcean maintains it has a high-growth business with $275 million annual recurring revenue.
Plenty of market share
For some alternative providers, specialization has the highest potential value.
Wasabi Technologies provides object storage services, founded in 2017 by serial entrepreneurs David Friend and Jeff Flowers, the co-founders behind backup software company Carbonite.
Wasabi offers general purpose cloud storage, designed to operate exactly like Amazon S3, except one-fifth the price, Friend, Wasabi co-founder and CEO, told CIO Dive.
To Wasabi, specializing in storage is a long-term value proposition. The company reports 13,000 customers with 5X growth in revenue in 2019 (Wasabi did not disclose the particulars of financials or how much data it stores). The company gets business from data-intensive industries, appealing to institutions such as universities and research labs.
"It's a very simple business model," said Friend. "We build out data centers, we fill them up with people's bits, and then we build some more data centers."
The company has five data centers in colocation facilities — three in the U.S., one in Europe and one in Japan — and has plans to double its infrastructure footprint this year.
The company is banking on an "other cloud" status. To Friend, the cloud race is akin to the mainframe business of the 1960s and '70s — every data center was dominated by IBM until the market began to diversify.
Amazon, Microsoft and Google threaten vendor lock in, Friend said. If a customer uses AWS, they also have to use its storage, computer, content delivery network … a laundry list of services but only with a single provider.
There's room in the market for companies to instead pick best of breed services and integrate across providers.
Ten years from now, most of the world's data will be in the cloud and Wasabi wants as much as it can get.
When the dust settles, a handful of companies will store the majority of the world's data, Friend said. "We want to be one of those."
For the few companies vying in a growing market, their long-term potential is lagging.
It's "not necessarily a winning strategy, but I'm not sure that any of these smaller companies in the market are thinking about being independent for the long run."
Research vice president, general manager at 451 Research
Operating as an anti-hyperscaler is a short-term value proposition, said Posey. There is room in the market for other providers, but not as standalone competitors.
These other companies survive with a certain amount of scale in place in order to support a hybrid, multicloud value proposition, Posey said. "The DigitalOceans and Linodes of the world, I'm not sure they do."
A lot of these clouds have medium- to long-term exit strategies for a sale, said Posey. Instead of hyperscalers, a potential buyer would be a professional services company or a global systems integrator where the cloud provider's tools would become one of many in a larger kit.
In the interim, companies, such as DigitalOcean, are focusing on individual developers or SMBs pick up on volume underserved by the hyperscalers.
It's "not necessarily a winning strategy, but I'm not sure that any of these smaller companies in the market are thinking about being independent for the long run," Posey said.