Cloud computing lets you use servers, storage, databases, and software over the internet instead of buying and running your own hardware. You can usually add or remove resources quickly and pay for what you use.

The most widely cited definition comes from Peter Mell and Timothy Grance at the US National Institute of Standards and Technology (NIST). It lists five essential characteristics:

  • On-demand self-service: you can set up resources yourself, without contacting the provider.

  • Broad network access: services are available over a network from many types of device.

  • Resource pooling: the provider serves many customers from shared resources.

  • Rapid elasticity: capacity can grow or shrink quickly as demand changes.

  • Measured service: usage is tracked, which supports pay-as-you-go billing.

The market is still growing quickly. In a September 2025 forecast, Gartner projected that public cloud services would grow 21.3% in 2026 and reach $1.48 trillion by 2029.

Types of cloud computing

Cloud services are usually grouped in two ways: by where they run and who can use them, and by how much the provider manages for you.

Deployment models: where the cloud runs

Model

What it means

Good fit for

Public cloud

A provider's infrastructure, open to many customers

Websites, apps, and changing workloads

Private cloud

Infrastructure used by one organization

Strict control or compliance needs

Community cloud

Infrastructure shared by organizations with common needs

Groups with shared security or legal requirements

Hybrid cloud

Two or more cloud types connected together

Moving gradually or keeping some data private

Many companies mix these models. In Flexera's 2026 State of the Cloud survey of about 750 cloud professionals, 73% said their organization runs a hybrid environment.

Service models: what the provider manages

Model

The provider manages

You manage

Examples

Infrastructure as a Service (IaaS)

Servers, storage, and networking

Operating systems, apps, and data

Amazon EC2, Google Compute Engine

Platform as a Service (PaaS)

Infrastructure and the operating platform

Your apps and data

Google App Engine, Heroku

Software as a Service (SaaS)

Almost everything, including the app

Your settings, users, and data

Gmail, Microsoft 365

A simple rule is that SaaS is ready to use, PaaS helps you build apps, and IaaS gives you the most control. Small businesses often coordinate that work with cloud-based project management tools.

Benefits of cloud computing

Team working together on laptops using cloud applications
Cloud tools let teams access their work from almost anywhere.
  • Lower upfront costs: you can avoid large hardware purchases and pay as an operating expense.

  • Easy scaling: you can add capacity for busy periods and reduce it afterward.

  • Faster launches: teams can set up new environments in minutes rather than waiting for equipment.

  • Global reach: large providers run data centers in many regions, which can bring services closer to users.

  • Remote access: teams can work from anywhere with a secure internet connection.

  • Stronger backup options: you can store copies in more than one location to support disaster recovery.

Risks to plan for

Person signing in securely to a business account on a laptop
Customers still need to protect their own data and access settings.

The cloud isn't automatically cheaper or more secure. These are the main issues to manage:

  • Costs: Flexera reports that 85% of respondents see managing cloud spend as a top challenge, and they estimate that 29% of their spend is wasted.

  • Security: providers don't handle everything. Under the AWS shared responsibility model, AWS protects the underlying infrastructure, while customers remain responsible for securing what they run in the cloud, including their data and access settings.

  • Outages: a provider failure can affect your service, so critical systems need a backup plan.

  • Vendor lock-in: moving between providers can be difficult if you rely heavily on one provider's tools.

  • Compliance: some industries and countries restrict where data can be stored and how it's handled.

Real-world uses of cloud computing

Streaming: Netflix

Netflix spent about seven years moving to AWS and completed its cloud migration in early 2016. Using AWS regions around the world lets it shift and expand capacity as demand changes. Netflix's video delivery also relies on its own content delivery network, so the cloud isn't its only infrastructure.

Banking: Capital One

Capital One closed its last data centers in 2020 after an eight-year effort and now runs its applications primarily on AWS. It had to rebuild thousands of applications and retrain staff. Along the way, a data breach exposed information from 106 million customers, a reminder that moving to the cloud doesn't remove security risk.

Healthcare: Mayo Clinic

A published case study describes Mayo Clinic's move to Google Cloud, including plans to migrate and anonymize data from more than 10 million patients. The platform also gives clinicians and researchers access to AI tools, with healthcare privacy controls in place.

Life sciences: Moderna

Moderna uses AWS to ingest, analyze, and store very large scientific datasets. AWS has also supported its research, development, and production work on mRNA medicines.

Everyday business

Most small businesses already use the cloud without thinking about it. Common examples include email, file sharing, accounting software, customer relationship management (CRM) tools, online stores, and video meetings.

How to get started

  1. Choose one workload. Start with something low-risk, such as file storage or a test environment.

  2. Pick the right service model. Use SaaS when a ready-made tool fits, and consider PaaS or IaaS when you need more control.

  3. Set a budget alert. Check your usage every month and switch off anything you don't need.

  4. Secure accounts early. Use multi-factor authentication, limit access, and encrypt sensitive data.

  5. Test your backups. Make sure you can restore your data before you need to.

Cloud computing works best when you match the service to a clear business need. Start small, track the cost and results, and expand once it's proving useful.

Sources

  • Mell, P., and Grance, T. The NIST Definition of Cloud Computing. NIST Special Publication 800-145, 2011.

  • Gartner. Forecast: Public Cloud Services, Worldwide, 2023-2029, 3Q25 Update. 2025.

  • Flexera. 2026 State of the Cloud Report press release. 2026.

  • Flexera. Cloud cost management trends. 2026.

  • Amazon Web Services. Shared responsibility.

  • Netflix. Completing the Netflix Cloud Migration. 2016.

  • SiliconANGLE. Capital One closes all data centers. 2020.

  • PubMed Central. Mayo Clinic's move to the cloud.

  • Amazon Web Services. Moderna customer story.