This functions in a way similar to utilities such as an electricity grid. A user just opens up a browser and accesses the required application and/or resource. The user then gets billed based on the usage, same as with utilities.
In order for this to work, providers offer data centers which simultaneously handle the resource requests from multiple clients. With everything in the data centers, the company does not need to have its own IT setup. This means companies can get rid of existing setups and the associated staff.
No doubt this saves a huge amount for the company, but there’s a lot more to it because the company can now avail of a data center’s massive capabilities. The resources it needs and pays for can be expanded or reduced whenever required. There will be no need to pay for upgrades or built-in excess capacity.
On the other side, providers have to be big companies with extensive technical know-how about managing a data center, not to mention very deep pockets. Any company that is planning to transfer its own IT setup to the cloud will naturally expect a high level of security, reliability and performance. Of course, a company can also do cloud computing in bits and pieces.
This means that some parts of the IT requirements can be fulfilled by a third-party who offers said requirement as a service via the internet. It can be IaaS or SaaS (Software as a Service) in which applications hosted elsewhere can be run on web browsers. Or it can be Platform as a Service, where a development environment is provided as a web based platform.