Public cloud vs private cloud: What’s the difference?

With a daunting dose of jargon and acronyms, the world of cloud computing can seem intimidating at first glance. If you’re wondering what the difference is between public cloud vs private cloud, you’re not alone. And you’re in the right place!

In this post, we’ll break down some of the basics of public cloud vs private cloud in a way that anyone can understand. We'll cover what public cloud is, what private cloud is, and the cloud architecture basics around both you need to know.

To start with, let’s give a quick definition of public and private cloud — a little TL;DR on the difference between public cloud vs private cloud.

• Public cloud is a term for cloud computing services delivered to multiple organizations over the public internet for free or pay-per-use.

• Private cloud is a term for cloud computing services dedicated to a single organization.

Ready to get a little deeper on public and private cloud? Read on as we take you from the basics up to… well, still the basics. (This is just a blog post after all.) If you're looking to go beyond the basics, browse our hands-on learning library. From manager to engineer and all roles in between, we have learning paths and IT skills assessments that can help you and your team get the cloud know-how needed to succeed.

Prefer to look at magical moving images rather than read words? For a solid (and totally free) primer on the basics of cloud computing, check out ACG’s Cloud Computing Foundations course. It’s one of the many courses free this month on ACG. 

• ELI5: What is the cloud?
• What is public cloud?
• What is private cloud?
• What is multi-cloud?
• What is hybrid cloud?
• What is the difference between public cloud vs private cloud?
• Why do companies pursue cloud adoption?
• Why do companies take a hybrid or multi-cloud approach?
• What does a cloud provider data center look like?
• What is virtualization?
• What is containerization?

Completely clueless as to this “cloud” word that gets tossed around all the time? Let’s start with an ELI5 (“explain it like I’m five”) primer on what the cloud is.

You can think of the cloud like electricity. Imagine you and your neighbors need electricity. (Hopefully not too hard to imagine given you’re reading this on something that runs on electricity.) 

You could produce your own electricity.

But you might find it easier to outsource the setup and operation of energy production to a third party that specializes in electricity production. By doing this, you and your neighbors can get easy, instant access to electricity with the flip of a switch. 

With this setup, you pay for what you use. Use more? Pay more. And vice versa.

The electricity company focuses on what it does well — getting you electricity quickly and efficiently and with minimal interruption — and you get to focus on doing your own thing.

Need that analogy decoded? 

• Creating your own electricity is like setting up on-premises computer resources (or, the old-school way of doing IT).

• Using a third party to create and manage your electricity would be like using a cloud provider. (As you might guess given the “cloud” in our name, we think this a pretty clever way to do things.)

• Paying for the electricity you use versus having to invest in infrastructure to generate your own electricity is like the capital expenditure (CapEx) versus operating expenditure (OpEx) discussion.
  
  What’s CapEx and OpEx? That’s a long story. But the CliffsNotes version is that moving from CapEx to OpEx means moving from investing in something you keep for years versus an ongoing business expense. (For an analogy on top of an analogy, that’s like paying to build up a library of MP3s versus paying a monthly subscription to Spotify.) Moving from CapEx to OpEx is one of the reasons many organizations migrate to the cloud.

Lift your teams to a common level of cloud knowledge! Scale your digital transformation learning faster with our accelerated Cloud Enablement Program.

When you hear people talking about cloud and cloud migration, it’s usually a safe bet they’re talking about public cloud. Public cloud deployments are the most common — not surprising given the many advantages of public cloud.

Public cloud is where services are offered over the public internet to anyone who wants to rent them. 

In a public cloud deployment, services and resources (like servers or storage) are owned and run by a third-party vendor over the public internet. Some of these services are free and others are pay-per-use.

These services are hosted by cloud providers like Amazon Web Services (AWS), Microsoft Azure, Google Cloud Platform (GCP), Alibaba Cloud, IBM Cloud, and Oracle Cloud.

These are called “public” clouds because they are open to the public. Anyone with an internet connection and a credit card can use their services. 

Some large traditional IT companies like IBM, SAP, and Oracle have created public cloud businesses to complement their traditional business. Microsoft Azure is a good example of this.

Private cloud is where computing resources are used exclusively by users from one business or organization. Companies may set up private clouds in their own data centers or with a hosting provider. 

With private cloud, cloud resources are used and owned solely by one organization. That makes this approach often desirable by governmental and financial industries that seek maximum control or customization

Private cloud providers include Hewlett Packard Enterprise (HPE), Dell, IBM, Oracle, and some familiar names from the public cloud provider space, including AWS, Google, and Microsoft.

Private cloud deployments can also sometimes be called on-premises deployments. Often, the private cloud deployment model is very similar to legacy IT approach, but it uses virtualization and application management to better utilize resources.

For another analogy to help drive the point home, consider how Nitya Narasimhan puts it in her Visual Introduction to Azure Fundamentals.

• Private cloud is like when a business owns and manages its own fleet of trucks. You have full control (data, security) and take on all responsibility (cost, maintenance).

• Public cloud is like when you rent trucks from a global company that everyone else uses as well. Resources may be shared or reused, and SLAs (service-level agreements) help you set up cost-security tradeoffs.

But back up a second. (Pun 100% intended.) What motivates businesses to adopt cloud technologies anyways?

Get the Cloud Dictionary of Pain
Speaking cloud doesn’t have to be hard. We analyzed millions of responses to ID the top concepts that trip people up. Grab this cloud guide for succinct definitions of some of the most painful cloud terms.

What are the benefits for businesses adopting cloud?

Businesses have been running like business as usual for decades. And much of what the cloud can do can be accomplished without the cloud. But the cloud introduces some new business capabilities and benefits.

Here are some motivations and benefits that lead many businesses to invest in cloud transformation.

• Fault tolerance — A common reason companies first get interested in the cloud is fault tolerance. If you’re a company that values your existence, you’ll have some sort of disaster recovery plan. These plans usually involve some alternate place where you store data or you’re able to recreate key business systems in the event that your primary data center has a problem of "oh no everything is broken, this is a fiasco" variety.
  
  Traditionally, this was done by contracting with a provider to keep a physical second copy of your hardware ready to go at a moment’s notice. Sounds good . . . but the problem is when everything is up and running fine, that backup physical hardware is just sitting around gathering dust and becoming obsolete. All while you’re still paying for it. Sure, it’s an insurance policy. But what if there was a better way?
  
  Rather than buying all your backup equipment, why not just rent it whenever you need it? Cloud providers have massive amounts of system capacity available to you in seconds. And you only have to pay for what you use, which is almost always a huge cost savings. (More on that in a second.)
  
  Then, when the crisis is over, you can just shut those things down and stop paying. That’s shifting from CapEx to OpEx. Remember that from above? It’s the same concept as taking an Uber or taxi versus buying a car. Now, you might say, "If I need that car all the time, then it would probably make sense over time just to buy the car." And you wouldn't be wrong, but in this case, if you’re constantly needing your backup site, you’ve got some much bigger problems on your hands than figuring out your transportation to dinner.

• Scalability — Another reason companies adopt the cloud is scalability. Scale is how much or how little capacity we need to meet business needs. Obviously, the goal is to have capacity as close to need as possible. But that’s a tricky thing to forecast. The pay-as-you-go model of cloud providers gives flexibility and the ability to scale up or scale down depending on our needs.

• Globalization — Globalization is another common reason for cloud adoption. As a company grows and expands beyond its home borders, it makes sense to have resources and services close to those new markets. This could be for regulatory or performance reasons. Well, good news! Cloud providers have data centers and resources in various geographies available, and you can use those with little more than a click of a button.

• Agility — Perhaps the most valuable business reason for cloud adoption is agility. Agility is the ability to respond to changing needs.
  
  In many companies, if you wanted to run an experiment that required IT equipment, you would likely have to endure a requisition and procurement process and get resources from the IT group to set up and maintain it. (That’s as fun and pain-free as it sounds.) And it could take weeks or months. In the cloud, you can get access to that equipment in minutes, try your experiment, and then shut down the equipment. You could get results in a day versus a month, it might only cost you a few cents, and you don't have to bug IT.

• Cost — Then there’s cost. Cost is a funny thing. (And not “funny ha-ha.”) While we would like to think cost is straight-forward, the reality is that most companies are not very good at tracking true costs.
  
  With cloud providers, costs are very literal. They show up on a bill. Some companies are surprised to learn that their costs might be more than what they believed their on-site data center costs to be. But pulling together the total cost of ownership (or TCO) is difficult. Are you including power, heating, cooling, and fire suppression in your calculations? How about security measures, like security cameras or card readers? Sometimes how your company chooses to account for your IT assets can significantly change what costs appear to be.
  
  At the opposite end of the spectrum, if you misuse your cloud resources, cloud can easily cost dramatically more than any sort of on-site data center. This is why it’s so important to train, reskill, and upskill your staff and have experienced cloud architects at the table.
  
  Cost savings can be realized with cloud, but be careful about the expectations you set — especially in the early days.

From the outside, cloud provider data centers look like big warehouses. Unless you look really closely, you probably couldn’t tell a data center from any other logistics warehouse. They’re usually heavily secured and have extra power and cooling equipment around them.

Inside there are rows and rows of racks holding physical computers. Each of these computers host several virtual systems (more on this below when we talk about virtualization) that get rented out to customers as they need them.

These data centers usually only have a small handful of people working there. Much of the maintenance is automated, and people are only needed when something needs a set of physical hands. 

Also, data centers are really, really loud. This is due to lots of cooling equipment running at full blast trying to keep everything from overheating.

Cloud ROI: How cloud skills generate real returns
We analyzed nearly 100 companies to determine the impact of a commitment to cloud maturity. In this guide, see how much value companies get when investing in cloud skills and technology.

There’s so much more to cloud than public and private cloud. And while the focus of this post is public vs private cloud, we should also touch on a few other concepts: multi-cloud, hybrid cloud, virtualization, and containerization.

Multi-cloud is the use of multiple cloud services in a single environment. This can mean a mix of public and private clouds or using multiple public cloud providers.

Many established companies will use a multi-cloud approach referred to as hybrid cloud. 

Hybrid cloud environments are the chocolate-vanilla swirl soft-serve combo of public and private cloud, combining aspects of both private and public cloud and sharing data and applications between them. 

Many companies opt for a hybrid cloud approach. And some companies also adopt a multi-cloud approach that involves using multiple public clouds. But why?

With hybrid cloud, it’s common for companies to use as much of their own computing resources as possible and then use public cloud resources to handle temporary exceptions.

Hybrid cloud can also be attractive to organizations that need to have some element of private cloud in place (for sensitive data or legacy IT infrastructure, for example) but still want to tap into the benefits of public cloud. The downsides? A hybrid cloud model can be more costly and complex, with the possible need for IT skills and on-premises hardware.

Some companies may opt for a multi-cloud strategy (like using AWS and Azure) to reduce reliance on a single provider, tap into benefits of more than one provider, or due to customer needs. (For example, imagine your business typically runs on AWS but you have Walmart as a customer, and they aren't super keen to give business to AWS because, you know, Amazon. You might need to be able to do work for them on Azure or GCP.)

Other people argue in favor of multi-cloud and using multiple public cloud to avoid vendor lock-in. Vendor lock-in is when you become so entrenched with a single vendor that you may lose flexibility or negotiating position.

However, with the current crop of public cloud providers, there’s not too much concern with vendor lock-in. There are ways to mitigate that risk. And, from a skillset standpoint, having to constantly pivot among multiple public clouds tends to create more challenges than it really helps.

Two other concepts you hear frequently around cloud computing are virtualization and containerization.

We talked earlier about how cloud providers have massive amounts of computing power available for you to use in seconds. Well, they don’t have people running around plugging in and unplugging cables whenever you request a new system. Cloud providers make use of virtual systems, or virtualization.

Virtualization means they have the ability to create software-based computer systems on top of physical computer hardware.

When you request a system, it can be set up — or “provisioned” — in a matter of seconds. When you’re done with it, it disappears, and the physical system resources can be reallocated to other customers. 

An evolution of virtualization is containerization.

To explain containerization, let’s look at the problem that containers are trying to solve.

Let’s say we have a brand-new computer system. Then, we start loading software on it. Sometimes one software package might create a conflict with another software package. This results in errors. Now let’s say we get a new computer system and we need to move over some software from that old system to the new system. And we accidentally leave out part of that software. This could result in problems too.

You may have experienced these problems on your own home computers, tablets, or mobile phones as you try to upgrade them. Imagine that problem across hundreds or thousands of systems and picture what kind of headache that would cause.

Enter: containers. Think of containers like a little backpack that holds everything you might need to travel anything you’d need to travel anywhere you want.

Containers are like virtualized systems but more lightweight. You can think of it as a nice, neat little box designed to keep things organized. Containers rely on something called a container engine that manages resources and allows the container to share those resources across themselves. So if one container isn’t working too hard, another container can take advantage of that. 

One of the most popular container engines is called Docker. 

Now, because containers are compartmentalized, they’re very portable. Since most major cloud providers also support containers, we can pretty seamlessly move those containers to a cloud provider and it will run just the same as it did originally elsewhere. 

Plus, we can use any one of those major cloud providers to host these applications. If we wanted to move the containers back on-site in a hybrid cloud model, we could do that too without much trouble. The portability of containers is one way companies are mitigating the vendor lock-in concerns we touched on above.

We’ve really just scratched the surface here in public cloud vs private cloud. Cloud architecture is a whole discipline that people spend years learning, so if you or your organization doesn’t have access to cloud architecture experts, seek the help of one before you design your cloud-based architecture. If you’re not careful, you can end up with quite a mess. 

Or, if you’re curious about how to become a cloud architect, check out this post.

Learn faster. Move faster. Transform now with courses and real hands-on labs in AWS, Microsoft Azure, Google Cloud, and beyond with our tech and IT development plans.