Cloud Computing 101: A Brief History of the Cloud and Its Evolution

Share:

Dear MarTech Maestro,

Can you break down the ever-ambiguous “cloud?” I know that iPhone photos save to the “cloud,” and most companies invest in “cloud solutions,” but what exactly is the “cloud?”

Sincerely,
CloudyWithAChanceOfSkepticism

A Sci-Fi Dystopia

CloudyWithAChanceOfSkepticism,

If you had told 10-year-old me that I would have a job working “in the cloud,” I’m not sure what kind of sci-fi dystopia I would have imagined – it sounds so ominous! What is the cloud? But before we can answer what the cloud is, it will be best to ask why the cloud exists. What made us create such a mysterious network of complexities in the first place?

I usually like explaining things through metaphor, but in this case, we can more directly compare to other types of resources. After all, technology infrastructure is not that different from physical infrastructure. Most companies need both to operate, and both have multiple dependencies like cost, size, and interaction with employees. It’s fun to tell people I’m an Architect and even more fun to tell them I am a “Cloud Architect” – they go from impressed to confused and skeptical faster than Elon Musk can lose $200 billion.

So in today’s blog, we will break down the 3 phases of the cloud’s evolution using office and employee management as an example. 

In phase one, companies bought buildings to house their employees. As companies had to keep up with maintenance costs and predict scalability, they moved to phase two; renting. In phase two, companies could outsource building management and employee resources. The modern working world was already on the move with the rise of WeWork and flexible spaces—then, wham, COVID-19 flipped everything on its head. Enter phase three, where remote work became the new norm, and thanks to Zoom and digital innovations, companies were no longer confined within the four walls of a physical building.

Just as companies need to allocate resources and spaces for employees, companies need appropriate physical spaces and accommodations for tech and servers. We can envision the evolution of tech into the cloud as the same kind of maximization of resources we’ve seen with the making of the modern office. Both providing an office and running tech require specific types of resources. 

PHASE 1: LET’S DO IT ALL IN HOUSE

In the early days of offices, a person or company would buy a building and everything in it. This came with a lot of upfront costs, and when things broke, or the company needed more space, they’d have to invest large sums of money again, only to eventually be stuck with too small of a building and outdated equipment. We’ll call this Phase 1 of physical architecture. These high costs for office investment and expansion meant you had to be rich or well-connected to even consider having your own company/office.

For officing, employees need a comfortable workplace that lends itself to productivity. Each employee should have a desk, a computer, a keyboard, and a mouse. Additionally, they need to be able to sit inside, protected from the elements, with enough space to work, and, ideally, some kind of climate control. All these amenities need to be maintained by office management and custodial staff. 

Likewise, Phase 1 of technology architecture was costly and inflexible. In the same way, businesses have to accommodate employees with appropriate resources in the office, every business needs to structure itself to deal with data storage and technology as well – a way to perform computations, store files, run applications, and connect to a network. For data and applications, you need a climate-controlled space, electricity, wires, cables, server racks, CPUs, memory, hard drives, software, an architected network, and someone who can run and maintain it all.

A company would buy a server or servers outright, where they would need to be physically configured and housed in a well-air-conditioned room near the workers they served. As a company grew or when newer technology was required, an entirely new server would need to be purchased outright, configured, and then all the data and programs would need to be migrated to it. The old server would be liquidated for pennies on the dollar because it was small, slow, and old. Meanwhile, the office air-conditioning bill continues to suck your hard-earned profit margin into the black hole, and the accounts payable department will no longer speak to you at all.

You can see in both cases, this model is very impractical and expensive. You likely need to purchase a server or building that is bigger than you need in case you grow, but if you go too big, that could be the very decision that prevents growth due to overspending on resources at the outset. 

Additionally, you need specialized labor to maintain everything – you are the landlord, and, in the case of tech architecture, you are also the only true line of defense against server crashes. Before high-speed internet was prevalent, this was the ONLY way to network all your people to a central system effectively. Phase 1 was expensive, time-consuming, and not particularly scalable or stable. And when someone decides to run the office microwave and dishwasher simultaneously, an ill-placed fuse could mean power loss, data loss, and, most importantly, sanity loss for the IT team.

PHASE 2: I’LL GLADLY PAY YOU TUESDAY FOR A NETWORK TODAY

Phase two came around quickly because phase one was so cumbersome and annoying. As broadband internet access connected networks together remotely (as opposed to actual wires running through the building like in Phase 1), server rooms could be located off-site. 

Building large server farms was more efficient than keeping a lone off-site server somewhere. These large warehouses popped up all over, and instead of housing servers on-site, companies would rent what is called a ‘bare metal’ server. These machines were basically the same racks of hardware as in Phase 1 but located elsewhere, using rented instead of purchased hardware. 

When your tech needs outgrew your server, you could call up the company and say, “Hey server farmer person, can you upgrade us from 256 megablobs to 512 gigablobs of computing juice?” and they would migrate all your stuff to a newer, bigger rack and charge you a little bit more per month. They would also maintain and replace all the hardware, so your IT department was free to focus on ping-pong programming and execution instead of wires and switches and blinky red lights.

This would be akin to renting a dedicated office in a commercial building in our physical architecture world. There may be other tenants, but they are not allowed outside their own part of the building. If a company needs to grow, it can find a newer, bigger office or even expand into adjacent space – if it is available. Costs are semi-predictable, and someone else fixes the things that break. If the company is unhappy with its environment or the market makes something better/cheaper available, it can make a change. It might be a slow change – leases, moving, contracts, and such all apply. However, it doesn’t suffer the decrease in equipment value; it just pays a little more for something a little better. The pricing is predictable, there are no upfront costs, and a certain amount of scalability is built in.

PHASE 3: HEAD IN THE CLOUDS

Bare metal servers still work fine for companies with predictable workloads, but growth and success are rarely predictable. What would happen if a business suddenly went viral and the employee count tripled? Chances are 2/3 of those employees would have a less-than-ideal working experience by pushing an office over capacity, immediately having to move offices, or splitting the workforce into multiple separate locations. 

Without digital advancements that allow virtual connectivity, employees working in separate buildings could make a larger company less productive and extremely chaotic. There’s an all-hands meeting and nowhere to have it! The new intern’s computer needs fixing, but the IT department is located elsewhere! Someone is at the reception desk for Ms. Important Big-Wig, but now she works 3 miles away! And the famous “I sent this to Printer 2a, and I might have sent it to an actual black hole.” What is really needed is infinite flexibility and far fewer ties to a physical location.

In 2020, the need for these things came crashing down around us as the COVID-19 pandemic made physical office spaces not just unscalable but actually dangerous. Companies had to change their facilities’ models overnight. The result was remote, flexible work. Each employee connected to everyone else remotely and worked from where they were. 

That was often their own kitchen table as the virus still hung heavy in the air, but eventually, things like coworking spaces also picked up steam. Individual employees could pay a daily or monthly fee for access to someplace like WeWork and enjoy any location they want at any time they want. If there was a big company push, everyone might be in a focused coworking space that day. And based on each employee’s individual needs for productivity, remote work allows for flexibility. The cost is directly related to the immediate need, not the estimated need. You pay for what you use and nothing more.

Granted, we had to develop not just new ways of thinking about work but also new working technologies. Chat apps, video meetings, and remote security policies have all changed to fit our needs. That is basically how the cloud operates. It’s essentially a server farm where instead of having specific pieces of machinery rented out to individual companies, all the computing power is pooled into a network of resources.

The cloud provider operates like a team of robotic facility managers who ensure everything runs smoothly. They set up walls around your data and computing when you need them for privacy, expanding and contracting as needed. They anticipate how much memory and computing power you need in real-time so that they can quickly grow and shrink the resources you are allocated. They swap hardware and clean up messes, ensuring that even if you don’t have the exact same server as before, it doesn’t matter because you have the same quality and experience – much like sitting at a different desk in a different location of a nationwide co-working brand. 

You only pay for what you use since pooling workloads in this mega-server environment tends to smooth out the usage spikes. If your server breaks, there’s a backup standing by that will slide into place without anyone noticing. It’s efficient, scales up and down, mitigates resource waste, and, thankfully, I don’t have to crawl around on the server room floor to make sure everything is plugged in at 3 am.

SO THEN, WHO IS BEHIND ALL THIS CLOUD BUSINESS?

Today there are three major cloud providers – Amazon Web Services, Azure by Microsoft, and Google Cloud. Amazon and Azure have the majority of the market share right now, but Google Cloud is making rapid gains. While Google is a newcomer in the public cloud arena, its tools have been in use for decades because Google Cloud was essentially developed as an internal platform to handle the needs of Google Search. 

At Adlucent, we use Google Cloud for many reasons, one of the big ones being easily connected Google Ads data. BigQuery, a Google Cloud-based warehousing and analytics tool, connects automatically to Google Ads data. Information can be shared, numbers can be crunched, and results can be visualized quickly and with extreme accuracy without ever leaving the Google Cloud interface.

At Adlucent, we have a remote workforce but also a physical office. I spend some days in-office, some days at home, and some days in the shiny new library down the street (which, by the way, is just as nice as any co-working space and is already paid for by my taxes. I like to refer to it as my ‘Socialism Office’). In technology, this is known as Hybrid Cloud – when you use a mix of on-premises servers and cloud servers depending on your needs at that moment. Likewise, you could be running on two different cloud providers – say, GCP and AWS. This is known as Multi-Cloud. 

CONCLUSION

In short, the cloud takes a hefty portion of the IT requirements off the shoulders of companies so that they can focus on business goals instead of IT infrastructure. It’s also far more full-service. Most cloud providers have hundreds of tools that are fully managed and integrated, making your applications easier to build, easier to scale, and blazing fast when you need them to be. So, when your photos are “In the cloud,” that generally means a few copies of them are sitting on a couple of different networked Linux servers in a warehouse somewhere. Those photos will meander around as servers come online and offline, floating here and there as if on a… well, a cloud.

I could write about the cloud – and Google Cloud specifically – every day, so I’m sure this isn’t the last you’ll hear about it. Send more cloud-based questions my way! In the meantime, make sure to subscribe to our performance marketing blog for more weekly insights like this.

Are you interested in discussing your MarTech needs? At Adlucent, we specialize in breaking down data silos to connect your first-party data sources and unlock customer and product-centric insights that fuel your business. Reach out today.

Who and What is the MarTech Maestro?

MartTech Maestro is here to help you navigate the ever-evolving world of digital marketing technology. With Blair Mundy, Adlucent’s Solution Architect, as your knowledgeable guide, you’ll never feel lost in a jungle of tech terms. The MarTech Maestro will break down complex tech jargon and answer the questions you’ve been too afraid to ask. Do you have a question for MarTech Maestro? Anonymity is guaranteed, so let your curiosity run wild!

Blair Mundy

Blair is Adlucent’s Solution Architect and holds the highly respected Google Cloud Professional Architect Certification. She’s been working in digital marketing and paid search for over 20 years and is particularly fond of pondering and predicting how technology changes the industry as a whole.