Platform as a Service (PaaS) provider Ep.io launched to rave reviews and developed a loyal fan base, finding a niche among Python developers as a "Heroku for Django" back before Heroku became the Heroku for Django by broadening its supported languages.
When Ep.io announced it was closing down its PaaS services in favor of a new plan to "fix servers," there was disappointment but little surprise: Cloud failures are common, and service providers need agility, particularly when a better established competitor attacks their "niche."
While Ep.io signed off, cryptically promising "exciting things in store," other cloud services providers have found ways to successfully leverage existing assets and strategies, jumpstarting new lines of business without having to reinvent the wheel.
Chute, a startup launched last year that offers managed media hosting through a simple API, was gaining traction among developers and drawing strong interest from large publishers. But the company ran into some recurring roadblocks when trying to actually close deals with larger companies.
A simplified solution to sell Chute's cloud offering
"It's very hard to offer [large customers] an API," said Ranvir Gujral, co-founder of Chute. "They need an SLA, and they need a much more productized offering." For a small startup still testing its hypothesis that cloud media management can offer compelling benefits far beyond simplified hosting, that could mean major investments in marketing, sales and infrastructure.
So Gujral and his team decided to follow another route: They would build a more complete, integrated offering on top of their cloud services, leveraging their existing infrastructure while better meeting the specific needs of their budding customer base.
The result was SlideChute, a drop-in, Flickr-like service that lets any publisher quickly accept and manage user-submitted photos. The service is still in beta, but the response has been fantastic, with big-name brands like NBC and Chevrolet already adopting it, Gujral said.
"What we realized is that you sometimes have to tailor your solution so that it becomes more usable," he said. "It helps unleash a wave of latent demand. People wanted to do this, but there wasn't a toolkit to do it."
Why couldn't Ep.io find cloud success?
Ep.io launched to a customer base hungry for Django-based PaaS before Heroku started offering a competing service. Ep.io co-founder Andrew Godwin said his company ran into a number of unexpected problems early on, but better timing of its own pivot might have made things much easier.
"We launched as Python-only at first, and by the time we'd added support for other runtimes it was a bit too late, thanks to our timing," he wrote in an email. "Had we launched that 6 months earlier I feel it would have had a lot less competition."
Godwin said that timing meant the company struggled to find a good market fit on its first go-around, compounding other difficulties -- ranging from thin margins to scaling -- inherent in the cloud services business.
Godwin said that Ep.io also struggled with pricing. Offering a "freemium" service, for example, helped garner early customers, but it also cut into the profitability of paying customers.
"Billing -- or, rather, coming up with a pricing plan that was both understandable and fair -- was an unexpected problem," he wrote, stating that multiple metrics of use such as RAM, bandwidth and CPU made simple plans difficult to make. "We had smaller margins […] since our free plan effectively reduced the amount each of the smaller customers was paying."
After cloud failures, a chance to reboot strategy
This sort of retreat-to-move-forward is not uncommon among cloud service providers, although the strategies are as varied as the types of clouds out there.
Huddle, which provides sync and collaboration tools, first tried a play as a freemium service pitched to consumers, before focusing solely on the more lucrative -- and complex -- requirements of the enterprise world, working hand-in-hand with IT rather than relying on rogue users for adoption.
FileTrek, formerly GridIron, actually pivoted toward a broader solution, from servicing the document management needs of the entertainment industry to focusing more broadly on file tracking.
Success with pivoting a cloud service's strategy has less to do with going niche or going broad, and more with tailoring the solution you can provide to real customers' needs, Chute's Gujral said.
"We really don't look at it as going niche. The market is so big: Even SlideChute is going after every brand and publisher," he said. "If you look at the model of companies that have come before us, they build a great product, they have a lot of developer activity, but to get customers, there needs to be a bit more of a sales and packaging of products."
In other words, while hotshot developers might help drive awareness and innovation with a service offering, they are not the ones who can write the big checks. The NBCs and Chevrolets of the world can help avert cloud failures by ensuring a steady supply of meaningful income.
Can cloud success be a waiting game?
Success is also a matter of timing.
While Dropbox and other consumer cloud-storage services capitalized on improved execution, the market's success also hinged on the timing of external factors, such as improved, pervasive Internet connectivity both at home and, on mobile devices, the increasingly sophisticated needs of consumers to access files from non-traditional devices and a rising comfort level with Software as a Service (SaaS).
Enterprises have been warier of rushing to the cloud, with some IT shops preferring to lean on tried-but-true technologies like VPN rather than handing data over to Google, Amazon or Microsoft, let alone a relatively unknown and untested cloud service provider. But that too might be changing.
Lynda Stadtmueller, a program director with Frost & Sullivan, has been researching the uptake in cloud services over the past few years. She has noticed an uptick in adoption with signs that the space is just heating up.
"We're going to see growth, and Storage as a Service is expected to grow faster," she said. "For Storage as a Service, we're predicting a 66.5% growth rate through 2015."
She said that Infrastructure as a Service, already significantly a larger market than storage, was primed to grow 52% through 2015. And Stadtmueller said that her numbers were conservative.
"If you talk to other analysts, my forecast numbers are lower than theirs," she said. "I'm seeing an absolute upward trend, but it's largely organic growth. We never saw what a lot of analyst firms predicted a few years ago, which is, 'Oh my gosh, everything is going to be cloud.'"
In other words, companies like Chute might avert cloud failures by embracing slightly less cloudy offerings, aiming for where the proverbial puck is going while making sure it can keep the lights on.
"Rackspace still earns a heck of a lot more from their traditional hosting services than their cloud services," said Stadtmueller. "Cloud gets all the press ... but it is not responsible for a big part of any major company's business plan."