What Is Cloud Computing?
Cloud computing lets you use powerful computers over the internet instead of owning them. Here's what it actually means, how it works, and why every company uses it.
It's 2 a.m. and your app just went viral
You built a small recipe-sharing app as a side project. A food influencer posted about it, and suddenly 50,000 people are trying to use it at the same time. Your laptop — which has been running the app from under your desk — catches fire. Not literally, but close. The website crashes. Everyone leaves. The moment is gone.
Now imagine a different version of this story. Instead of running the app on your own machine, you set it up on someone else's computers — massive, powerful ones sitting in a data centre in Virginia. When 50,000 people show up, those computers notice the spike and automatically spin up more capacity. Your app stays online. The influencer's followers sign up. You wake up to 50,000 new users and a $4.80 bill for the extra computing power you used overnight.
That's cloud computing. You didn't buy a bigger computer. You didn't call anyone. The cloud just handled it.
So what is "the cloud," really?
Strip away the buzzwords, and cloud computing means this: using someone else's computers over the internet instead of buying and maintaining your own.
That's it. When someone says "it's in the cloud," they mean the software, the files, or the processing power lives on computers owned by companies like Amazon, Microsoft, or Google — not on a machine sitting in your office.
You already use cloud computing every day, even if you've never thought about it:
- Gmail — your emails aren't stored on your phone. They're on Google's servers.
- Netflix — movies aren't downloaded to your TV. They stream from AWS data centres.
- Spotify — your playlists live on Spotify's cloud infrastructure.
- Dropbox — your files sync across devices because they're stored in the cloud.
- Google Docs — you edit documents in a browser, saved automatically to Google's servers.
If your laptop died right now, you could log into any of these services from a new device and everything would still be there. That's the cloud at work.
Before the cloud: the old way
To understand why cloud computing matters, you need to understand what came before it.
✗ Without AI
- ✗Buy servers for $50K–$500K upfront
- ✗Hire IT staff to maintain hardware
- ✗Guess how much capacity you'll need
- ✗Wait weeks to set up new servers
- ✗Pay for idle capacity 24/7
- ✗Hardware becomes obsolete in 3–5 years
✓ With AI
- ✓Pay only for what you use, monthly
- ✓Provider handles all hardware maintenance
- ✓Scale up or down instantly
- ✓Launch new servers in minutes
- ✓Shut down resources when not needed
- ✓Always running on current hardware
Companies used to buy servers, put them in a room (or a closet), wire them up, install software, and pray nothing broke on a Friday night. This is called on-premises (or "on-prem") infrastructure. Some companies still do this — and for certain use cases, it makes sense. But for most businesses, the cloud is cheaper, faster, and less stressful.
There Are No Dumb Questions
"If the cloud is just someone else's computer, why not just buy my own computer?"
You can! But think about electricity. You could install a generator in your backyard and produce your own electricity. Instead, you pay the electric company a monthly bill because they can produce electricity at massive scale for a fraction of what it would cost you alone. Cloud computing works the same way — Amazon, Microsoft, and Google buy millions of servers, run them more efficiently than you ever could, and rent out small slices to you.
"Is the cloud less secure than keeping things on my own computer?"
Usually the opposite. Cloud providers employ thousands of security engineers and spend billions on security infrastructure. Your office server protected by a simple password is almost certainly less secure than AWS. That said, misconfiguring cloud security settings is a real and common problem — so the cloud is only as secure as the person setting it up.
The three flavours of cloud: IaaS, PaaS, SaaS
Not all cloud services are the same. Think of it like housing:
| Model | What it stands for | What you get | Housing analogy | Example |
|---|---|---|---|---|
| IaaS | Infrastructure as a Service | Raw computing power, storage, networking — you build everything on top | Renting land — you get the plot, but you build the house, install plumbing, paint the walls | AWS EC2, Google Compute Engine, Azure VMs |
| PaaS | Platform as a Service | A ready-made environment to build and run applications — no server management | Renting an apartment — walls, plumbing, and electricity are handled; you just bring your furniture | Heroku, Google App Engine, Cloudflare Workers |
| SaaS | Software as a Service | A finished application you use through a browser — no building required | Staying at a hotel — everything is done for you; just show up and use it | Gmail, Slack, Salesforce, Notion |
Most people interact with SaaS every day without realizing it. Developers and engineers work more with IaaS and PaaS. If you're in a non-technical role, you'll mostly care about SaaS — but understanding all three helps you speak the language when your engineering team says "we're migrating to IaaS."
Name That Cloud Model
25 XP2. A startup deploys its new app on Heroku without worrying about server configuration. →
Public, private, and hybrid cloud
Besides the service model (IaaS/PaaS/SaaS), clouds also differ in who can access them:
| Type | What it means | Who uses it | Example |
|---|---|---|---|
| Public cloud | Shared infrastructure, open to anyone who pays | Most companies, startups, individuals | AWS, Azure, Google Cloud |
| Private cloud | Dedicated infrastructure for one organisation | Banks, governments, healthcare | A hospital running its own cloud in its own data centre |
| Hybrid cloud | Mix of public and private — sensitive data stays private, everything else goes public | Large enterprises with compliance requirements | A bank using private cloud for customer data but public cloud for its marketing website |
Most companies use public cloud — it's the cheapest and most flexible option. Private and hybrid clouds exist for industries where regulation demands tighter control over data (think medical records, financial data, classified government information).
The big three cloud providers
Three companies dominate the cloud market:
| Provider | Market share (~2024) | Known for | Famous customers |
|---|---|---|---|
| AWS (Amazon Web Services) | ~31% | Broadest service catalogue, first mover | Netflix, Airbnb, NASA |
| Microsoft Azure | ~25% | Enterprise integration, Office 365 synergy | Walmart, Samsung, the US government |
| Google Cloud (GCP) | ~12% (Synergy Research Group, Q4 2024) | Data analytics, AI/ML, Kubernetes | Spotify, Twitter/X, PayPal |
Market share figures per Synergy Research Group / Statista, Q4 2024.
Together they control about two-thirds of the global cloud market. Other players include Alibaba Cloud (big in Asia), Oracle Cloud, and IBM Cloud.
Why companies move to the cloud
The risks (because nothing is free)
Cloud computing isn't perfect. Here's what can go wrong:
- Internet dependency — No internet, no cloud. If your connection goes down, your entire business could stop. (This is why redundant internet connections matter.)
- Data privacy — Your data sits on someone else's servers. If you're a European hospital, storing patient data on US servers may violate GDPR. You need to know where your data lives.
- Vendor lock-in — As mentioned above, migrating away from a cloud provider is expensive and time-consuming once you're deeply integrated.
- Ongoing costs — Cloud is cheap to start, but costs can spiral if you're not monitoring usage. Many companies have been shocked by unexpected cloud bills.
- Outages — Even the biggest providers go down. When AWS has an outage, half the internet notices because so many services depend on it.
There Are No Dumb Questions
"Can I use multiple cloud providers at once?"
Yes — this is called a multi-cloud strategy, and many large companies do it. They might run their main application on AWS, use Google Cloud for data analytics, and keep Azure for Office 365 integration. It reduces vendor lock-in but adds complexity.
"Is cloud computing the same as web hosting?"
Web hosting is a tiny slice of cloud computing — it's like asking if a bicycle is the same as transportation. Web hosting gives you a place to put a website. Cloud computing gives you entire operating systems, databases, AI training platforms, IoT management, and thousands of other services. Every web host today is probably running on a cloud provider behind the scenes.
Cloud Benefits vs. Risks
25 XP2. A company's cloud bill triples unexpectedly after a traffic spike. →
Cloud vs. on-premises: when each makes sense
Cloud isn't always the answer. Here's a practical guide:
| Factor | Choose cloud when... | Choose on-premises when... |
|---|---|---|
| Budget | You want low upfront costs and monthly payments | You have capital to invest and want predictable long-term costs |
| Scale | Your needs fluctuate (seasonal traffic, rapid growth) | Your workload is steady and predictable |
| Compliance | Your data doesn't have strict residency requirements | Regulations require data to stay in specific physical locations you control |
| Team | You don't have (or want) a large IT operations team | You have skilled IT staff who can manage hardware |
| Speed | You need to move fast — new projects, experiments, MVPs | You've already built your infrastructure and it works |
Most modern companies use a mix: cloud for most things, on-premises (or private cloud) for the sensitive stuff.
Why "cloud experience" is on every job posting
Open any job listing in tech — or increasingly in marketing, finance, HR, and operations — and you'll see "cloud experience" listed as a requirement or a plus. Here's why:
Every department now uses cloud-based tools. Marketers use HubSpot (SaaS). Designers use Figma (SaaS). Salespeople use Salesforce (SaaS). HR teams use Workday (SaaS). Finance teams use cloud-based ERP systems.
Understanding cloud computing isn't just for engineers anymore. It's baseline literacy for working in a modern organisation — the same way knowing how to use email became non-negotiable in the 2000s.
Cloud Computing in Your Life
50 XPBack to the 2 a.m. viral spike
In the fixed-infrastructure version of the story, the recipe app crashes under 50,000 simultaneous visitors and the moment is gone — the influencer's audience bounces, the opportunity evaporates, and you wake up to a dead server and zero new users. In the cloud version, the spike is handled automatically: compute capacity scales up within seconds, the app stays online, and you wake up to 50,000 new sign-ups and a hosting bill that cost less than dinner. The difference wasn't the code or the product — it was whether the infrastructure could respond to demand in real time. That's what cloud auto-scaling delivers: the ability to absorb unpredictable growth without a 2 a.m. phone call, a panicked server purchase, or a missed opportunity that never comes back.
Key takeaways
- Cloud computing = using someone else's computers over the internet. You rent computing power instead of buying it, the same way you rent electricity instead of running a generator.
- Three service models: IaaS (you manage everything), PaaS (you manage the app), SaaS (you just use it). Most people use SaaS daily without realizing it.
- Three deployment types: public (shared, cheapest), private (dedicated, most control), hybrid (mix of both).
- AWS, Azure, and Google Cloud control about two-thirds of the market (~2024, Synergy Research Group).
- Benefits: scalability, cost flexibility, global access, automatic updates, disaster recovery.
- Risks: internet dependency, data privacy concerns, vendor lock-in, cost surprises, outages.
- Cloud literacy is a baseline professional skill — not just for engineers, but for everyone working in a modern organisation.
Knowledge Check
1.A startup needs to launch an app quickly without buying any servers. They want to focus on writing code, not managing infrastructure. Which cloud service model is the best fit?
2.A hospital needs to store patient records digitally. Regulations require that patient data stays within the country and on infrastructure the hospital controls. Which cloud deployment model best fits this requirement?
3.Which of the following is a real risk of cloud computing that companies should plan for?
4.You use Gmail for email, Slack for team chat, and Google Docs for writing — all through your web browser. What do these three services have in common?