The year was 2008 and the Interop tech trade show in New York City was crammed full of booths. Sales reps offered trinkets as they hawked their next-gen software and hardware. I wondered through blinking displays and the noise of a thousand buzzwords.
The brand name vendors – Microsoft, Intel, Oracle, IBM – had paid big bucks for booths the size of small houses. Large crews in company shirts scanned your badge for a later marketing push. Staffers gave product lectures backed by full-size video screens.
After touring these big outfits, I investigated the smaller booths hosted by mid-sized players. With minimal staff, they worked still harder to lure you to their pitch. In an era before “Booth Babes” were outlawed, some booths included twenty-something women in skimpy sequined outfits, handing out t-shirts or glow-in-the-dark key chains.
Feeling dazed, I wandered into the still more low key area. Tiny booths staffed mostly by bare bones crews. There I saw a modest booth by an outfit called Amazon Web Services. A sole rep manned it, and he wasn’t wearing a company shirt. He had no giveaways or promotional games.
No one was talking to the guy, his booth at that moment was completely ignored.
I had heard of AWS, but – I’m embarrassed to admit this now – I didn’t think it was a big deal. But since I didn’t have to wait in line, I thought, What the heck, I’ll check out his story.
Funny thing about tech trade shows: the vendors are terrified of being videoed. Even if you’re a tech journalist simply videoing their sales pitch (as I was), it must be approved by three layers of management. Merely asking to video earns you anxious looks.
But when I asked the AWS guy if I could video his pitch, he said “sure” – completely unconcerned, no red tape. He spoke casually. He spiel was so modest that he actually used the phrase “very primitive” in describing AWS. (See video below.)
Yet the message that young guy shared that day, alone in his lowly booth, would upset the entire $3 trillion global IT industry. Vast sums would be spent resisting it. Then, slowly, truly gargantuan sums would be spent accepting it.
Finally, the technology that the young rep spoke of would be embraced wholeheartedly, and in the process would fundamentally change modern life.
World, meet cloud computing.
At the 2008 Interop trade show in New York City, an AWS cloud rep described his company’s cloud platform as “very primitive.” (1:48)
Cloud Starts to Change the World: “Confusion Will Remain”
Computing has traditionally been about clunky boxes that contain metal and silicon. A desk in your home is likely decorated by a clunky PC that holds metal and silicon. Large companies have entire rooms – datacenters – full of expensive boxes that hold silicon and metal.
These boxes are marvelous things, but they need a lot of love. They require electricity and software updates. They crash regularly. They arrive at the datacenter and highly paid IT pros have to hook them into the network.
And the very moment they’re hooked up? They start hurtling toward obsolescence.
Cloud computing wipes that world away. With cloud, you rent your clunky box of metal and silicon. The rental agency – a cloud company – is remote. Its limitless computing resources are floating off in a distant reserve.
The cloud company pays the electric bill. They hire IT staff to babysit the machines. When the boxes get outdated, they upgrade them. They worry – obsessively – about security.
All you do is log-in remotely. (And send the cloud company a monthly check, of course.)
Great idea, right?
Actually, most businesses in the early years of cloud looked at it and shrugged. Not interested. Gartner, in a 2008 research report, noted that “the concept of the cloud has emerged and is now gaining traction.” However, the research firm opined sagely, “confusion will remain throughout 2008.”
Indeed, confusion would remain. Businesses said, We have a data center. We’ve spent a king’s ransom on it – we don’t need yours.
Furthermore, they kvetched: Our computers hold the guts of our business. You want our precious data to flow outside the firewall? What about security? What about compliance? What about hiring staff that understands it?
Maybe later, they said – we’ll think about it.
Ellen Rubin, CEO and Co-Founder, ClearSky Data: “There’s now a whole generation of startups that were able to get going without needing to do all the things they needed to do in the past.” (1:11)
Cloud Before the Cloud
In truth, even when AWS launched in 2006, cloud computing wasn’t as new as it seemed. Salesforce.com, founded in 1999, had turned Software as a Service (SaaS) into a mainstream idea. SaaS can be thought of as “cloud computing lite.”
With SaaS, users rent software from a remote provider. The vendor handles the headaches of managing the hardware and infrastructure to support the software. All users do is log in and use the software.
Like cloud itself, SaaS started slowly. The Salesforce IPO in 2004 raised a comparatively modest $110 million. (In contrast, Zoom, a SaaS company that went public in 2019, raised $14.4 billion).
Users in the pre-cloud, pre-SaaS era bought “shrinkwrapped” software. It actually came shrinkwrapped in attractive boxes, and you installed it manually on your machine. SaaS ran a stake through the heart of packaged software.
To many companies in the early 2000’s, SaaS was far less threatening than full-fledged cloud. SaaS is merely about renting something you once bought. In contrast, cloud entails accepting that your computing platform will be off premise. For many companies, that presented a much thornier challenge.
Sam Charrington, Founder, TWIML: “Cloud enabled a lot of the growth that we’re seeing in AI today.” (1:35)
Cloud vs. the Business World’s Big Secret
Resistance to cloud computing was weakened by a barely concealed secret: companies despise their datacenters.
Big hulking datacenters are seen as a necessary evil. No one wants to build one. They’re certainly impressive – they resemble the set of a science fiction movie, row upon row of silent machines. But they’re wildly expensive, and their maintenance challenges are never ending.
Yet if you’re a market-dominating vendor – or if you hope to be – you need an in-house computing behemoth. Or so the thinking went.
Before the cloud, owning a datacenter did separate the major players from the also-rans. Among other advantages, in-house computing facilities provide whiz-bang data crunching tools that (in theory) offer deep business insight.
So companies grudgingly opened their wallets and showered mountains of dollars on these computing Tyrannosaurus Rexes. And these facilities were huge.
A company called Switch, for instance, opened a 1.5 million square foot datacenter in Nevada. On opening day, the ribbon was cut by a robot nicknamed Rebel Metal, a humanoid-looking creature that can drive a car and walk up steps.
Even for companies that openly doubted their datacenters (“maybe we move toward this cloud thing?”) inertia was powerful. Alas, breaking up is hard to do. Making it harder: the IT industry kept trying to make the datacenter seem hip and futuristic.
For instance, how about beefing up your datacenter with nifty new hyperconverged infrastructure? HCI combines compute, storage, and networking into one seamless system, ostensibly easing life for datacenter administrators.
It’s pretty snazzy, until your realize that an HCI system is a hunk of metal and silicon that needs electricity and care from expensive staff. It gets old and – unlike the compute power rented from the cloud – is a depreciating asset.
The real magic of cloud computing is its limitless scalability. In Star Trek, when Captain Kirk yells, “Scotty, we need more power!” more power appears pronto. Likewise with cloud: click on your dashboard and AWS (or Azure or Google Cloud) provides a computing surge.
Trying doing that with an HCI system. Or with an entire datacenter.
Another factor supporting the datacenter in cloud’s early years: IT staff who made healthy salaries babysitting data centers. They could too clearly see the advantage of cloud computing. But if management ever figures it out (which would happen with glacial slowness), what about our great in-house IT jobs?
Unfortunately, the move to cloud did cause job losses. In 2015, gaming firm Zynga axed some 360 staffers when it migrated features to AWS. Microsoft in 2017 cut thousands of jobs as it pivoted to cloud sales. Of course with the tech industry’s hunger for talent, many IT professionals can find new jobs. But this may require some scrambling – it’s easier said than done.
JP Gownder, VP, Principal Analyst, Forrester: Cloud is “a hyperscale extension of fundamental business processes that solves problems, anywhere in the world, any time, at scale.” (2:16)
No, the World Won’t Change: Creative Resistance to Cloud
Diane Greene is little known outside the arcane world of enterprise IT. But within this world, Greene is a major player who shaped tech history.
In 1998, armed with two graduate degrees (one in computer science) and experience in startups, she formed a company with her husband, Mendel Rosenblum, a professor at Stanford, and a small group of tech talents. They called their budding venture VMware. The fledgling company had some 20 employees.
They grew fast. With Greene as CEO, VMware made virtualizaton the foundation of most every datacenter in existence, worldwide.
Virtualization places a layer of software between the hardware of computer servers and the software that runs on those servers. This transforms a datacenter full of clunky metal servers into a unified computing environment. Virtual machines can be spun up at a moment’s notice. A virtualized datacenter is more efficient and more automated, and spends less on hardware.
Companies loved it. Greene and her crew soared upward. In 2003, industry giant EMC bought VMware for $625 million. By 2008, when Greene was pushed out by EMC head Joe Tucci (in what some viewed as a questionable decision) VMware’s market cap had rocketed to north of $20 billion.
Yet something funny happened when VMware’s starry ascent ran into the early years of cloud computing.
The collision created an odd idea: private cloud computing.
Companies, having virtualized their datacenters, now called this unified compute pool a “private cloud.” Which was strange. The thinking seemed to be, “Hey, we’re on board with this new-fangled cloud thing – in our very own legacy datacenters.”
Private cloud suffers an obvious limitation: it’s still an in-house datacenter. It’s fancier and does far more tricks, but it lacks the power and scalability of a remote cloud deployment. A company still has to buy clunky boxes that tumble into obsolescence.
Critics, to be sure, poked fun at private cloud, opining that it was just a dressed up old-fashioned data center. In about 2010, the term “cloud washing” was born. It referred to the vendor practice of adding the term “cloud” to any traditional tech product. IT vendors of every vintage rewrote their product pages to sell their legacy products as “cloud focused.”
Still, private cloud did produce change: it led the way to hybrid cloud. With hybrid cloud, companies bridged their in-house datacenter – their private “cloud” – with an actual public cloud. This combination allowed a company to leverage cloud’s scalability even as it clung the comfort of its traditional data center.
Resistance, clearly, was fading.
Judith Hurwitz, President, Hurwitz & Associates: “The greatest impact [of cloud] is really on speed.” (2:26)
The World Shifts: Cloud’s Early Pioneers
While many businesses looked warily at cloud, Amazon Web Services found some early enthusiasts.
Startups were thrilled. Before cloud, only deep-pocketed companies had datacenters. With AWS, startups simply brandished a credit card and – presto – started leveraging powerful computing tools. Legions of entrepreneurs with big dreams but little cash could now compete. Cloud became the great business equalizer.
Likewise for software developers. AWS’s nascent cloud opened the door to a giant toy store. All the storage and compute tools you need – Yippee! – and no need for approval by management bean counters. Countless app projects took flight.
Another early AWS customer group was – ironically – staff at companies that ignored cloud. In a phenomenon is known as shadow IT, employees use the departmental credit card to launch cloud accounts – without management approval. Theses staffers needed more computing resources than their company offered.
Shadow IT is “rogue computing” from the company’s standpoint. One day the firm’s accountant discovers that – gasp! – there’s 8 new cloud accounts. Did someone authorize this?
Astoundingly, Gartner in 2019 reports that about a third of corporate IT spending is shadow IT. The prevalence of unsanctioned cloud accounts reveals that rent-on-demand compute tools are too useful to resist.
Also, companies understand (without acknowledging it) that it’s cheaper to outsource a tech platform than buy in-house resources.
The true world-changing effect of shadow IT: staff at companies of all sizes are empowered by an expanding toolset. Have cloud, will travel.
Jim Hare, Research Vice President, Gartner: “A lot of the new businesses that have emerged – they’re all built on the cloud.” (2:10)
The Cloud Battle – and the World’s Biggest Toolbox
As cloud computing grew, the top enterprise IT vendors faced an embarrassing truth: they had been upstaged by Amazon – previously known for selling used books and microwave popcorn makers.
In fact, not just upstaged. Legacy tech vendors had watched as their industry had been (to use the buzzword) disrupted. And actually, not merely disrupted – Uber disrupted taxi service but still uses the time-honored model: users request a car and get ferried across town. Same model, just better executed.
What Amazon did by displacing the in-house datacenter was offer a completely different model. The off-premise cloud platform reshaped the very foundation of how technology powers the world.
It was as if every tech vendors had been in, say, the taxi business. But one vendor offered to helicopter customers to their location; the flexibility and power of cloud is that much of a quantum leap.
Acknowledging this terrible truth was painful and slow. In 2008, Oracle CEO Larry Ellison disparaged cloud as an overhyped fad. “When is this idiocy going to stop?” he implored.
Ellison voiced the angst that many vendors were feeling. Selling gear for the data center – that bottomless money pit – was a gushing fountain of revenue. Selling old-fashioned packaged software had been a license to print money.
Slowly, the legacy tech vendors realized: we need to get on board! The train is leaving the station. And thus over a few years, the Great Cloud Battle began: AWS vs. Azure vs. Google Cloud.
Playing catch up, Microsoft launched its cloud platform in 2010, an agonizingly long four years after AWS’s debut. Its original name, Microsoft Windows Azure, reflected the company’s roots in packaged software.
When Satya Nadella became Microsoft CEO in 2014, he assertively shifted the company toward the cloud. He announced a “cloud first, mobile first” strategy. In a titanic shift, the company removed “Windows” from the name of its cloud, simply calling it “Microsoft Azure.” In the cloud, Windows isn’t the center of the world anymore, even for Microsoft.
Microsoft played a smart game of catch up, using its historic strength in the enterprise to romance cloud customers. The company embraced the hybrid cloud, which is popular among large companies that can only move so fast. Over a few years, Microsoft pulled to a respectable second place behind AWS. (And Microsoft’s stock price, after languishing under CEO Steve Ballmer, zoomed upward nearly 300 percent with Nadella’s leadership.)
Google, with its gargantuan network, was perfectly positioned to compete in the cloud, yet struggled with strategy. After an early, ill-considered effort, Google Cloud Platform didn’t formally launch until 2013. Despite deep strength in data analytics and AI, Google lacked a connection with the enterprise buyers. Google has so far been a perennial laggard behind cloud leaders AWS and Azure.
Google wasn’t the only cloud vendor playing catch up with AWS’s first-mover advantage. IBM, Alibaba, Oracle, VMware, Dell, Rackspace and a host of others invested jaw-dropping sums in developing cloud offerings, in various models.
The point: these leading vendors – really, the entire IT sector – acknowledged the old world of the datacenter was receding. The new, emerging world of cloud computing was now dominant. Each vendor realized: we either embrace the cloud wholeheartedly or risk extinction.
Percentage of enterprises that will close their traditional datacenters in 2019, and in 2025. (Source: Gartner, February 2019)
Andres Rodriguez, CTO and Founder, Nasuni: “The premise of [cloud] architecture is that everything can be distributed, and the distributed system can scale forever.” (:59)
The Cloud Spins Faster and Faster
Two factors combined to drive a massive uptrend in cloud computing between roughly 2010 and 2015 – a shift that would greatly amplify cloud’s effect on the world.
First, the growing crowd of vendors jumping into the market drove intense competition. Each rushed to offer more and better cloud solutions; a price war ensued, which helped drive adoption of cloud.
Second, and related: this period saw a dazzling rush of tech innovation. A dizzying array of new technologies emerged.
All these emerging technologies either sprung from the cloud or closely interoperated with cloud: serverless, containers, Kubernetes, microservices, blockchain, DevOps, IoT, machine learning. Cloud computing morphed into multicloud, a model in which companies used two (or more) cloud platforms to access the best from each.
The net effect of all these advances: Where once there had been a single, staid entity, datacenter, now there emerged a dynamic platform – cloud – that prompted a full flowering of ever more powerful systems.
Making cloud more central: Many companies looked to the cloud to purchase new technologies. Companies could never build all these whirly-gig software tools and systems by themselves; it would be too complex and expensive. But the cloud vendors offer them on an affordable pay-as-you-go basis.
This is particularly true in the critical areas of artificial intelligence, machine learning, and data analytics – anything that requires algorithms and intense data processing.
As much as any factor, it’s the cloud that enables the power of AI and analytics to be widely available. Each of the major cloud companies offer robust AI solutions. And an entire legion of smaller and mid-size cloud-based vendors also offer AI solutions.
Yet as important as AI and data tools are, they are just a small fraction of the available cloud-based tools.
The earlier promise of Software as Service has came to fruition: everything is now offered as service, including:
- Backend as a Service (BaaS)
- Database as a service (DBaaS)
- Machine Learning as a Service (MLaaS)
- Disaster Recovery as a Service (DRaaS)
- Blockchain as a Service (BaaS)
And the list is growing even as you read this.
Even leading retail software programs have made the switch. In 2013, Adobe announced that new releases of its Creative suite (Photoshop, Premiere) would only be available via the cloud, on a SaaS subscription basis.
Users howled in protest – they loved the “buy once and install” model. But Abobe stood firm. The software is more easily updated via the SaaS model; the lower monthly fee attracts a wider audience than an expensive one-time purchase. Most important to Adobe: moving to a SaaS-only model produced record revenues for the company.
Microsoft Office has been the ultimate shrink-wrapped “buy once and install” cash cow. Sales of the productivity suite were central in establishing Microsoft as the reigning software giant.
In 2011, Microsoft debuted a SaaS version, Office 365. Unlike Adobe, Microsoft continues to sell shrink-wrapped versions. In 2017, a historic shift took place: revenue from Office 365 sales topped revenue from the traditional shrink-wrapped version. If there was a single signpost denoting that the dominant computing model had shifted, this was it.
The Cloud Era had arrived.
Krishnan Nandabalan, CEO and Founder, InveniAI: “Cloud computing has democratized the ability to use data.” (1:20)
How Cloud Computing Changed the World (But a Question Lingers….)
As the year 2020 looms – the date itself seems futuristic – cloud computing’s effect on the world has been profound: it has dissolved the boundary between humans and technology.
When the datacenter was king, the magic computing tools were held behind high walls. Only a select few could access the castle with the hyper-powerful tools.
But the king has been overthrown. As computing shifted from the datacenter to the cloud, the high walls have vanished. In the new world, the unwashed masses, everyone (or everyone with a few bucks) can access the magic tools. These tools – data, AI, development, business, media, health, entertainment – are now omnipresent.
Cloud makes technology immersive, for better or worse. There is, in effect, no longer any boundary between we humans and an ever expanding computing environment. It’s at our finger tips, in our ear buds, on our screen – all of our screens, everywhere, all the time. It shapes the very fabric of the world we now live in.
To be sure, cloud-based technology is enabled by delivery systems: the Internet and mobile devices. But the Web and mobile are just that – delivery mechanisms. The Web itself can’t perform, say, machine learning or artificial intelligence. Your phone itself (though wildly over-priced) is just a chunk of metal and silicon.
It is cloud computing platforms – accessed via the Web and mobile – that flip the switch on the world. That shifts it from a world where technology must be reached for into a world where technology is virtually the air we breathe.
Countless transactions are enabled by cloud. A woman riding a yak in Nepal tracks her business sales with an analytics app on her mobile device. A small company with dreams of improving healthcare spins up an AWS account and leverages databases. Two people across the planet from one another use a video streaming app to share a logistics modeling program. A high school student accesses an AI system and learns skills for a later career.
Healthcare systems, the financial world, media companies, educational institutions – all are powered in part (or largely) by cloud computing platforms. The modern world, in all its science fiction reality, is enabled by an oceanic network of cloud-based connections.
All wonderful, perhaps, yet this new human epoch of cloud-based empowerment creates a major question. When most everyone on the planet (in theory) has access to august computing power, what will be the result? How will that change us? In the years ahead, to what end will we employ the awesome power that cloud has made so easily accessible?