Before the cloud computing model, IT departments had a long, expensive, and hard path to follow to make a solution available to its customers. They had to buy servers; wait for delivery; provision space in their data centers; prepare networking and virtualization; install operation systems, middleware, databases, and runtimes; and then develop applications and deploy them to the production environment.

When cloud computing takes place in the same scenario as presented above, provisioning the same computing capabilities and start developing applications happens in a matter of minutes. You will only need to pay for the use of the resource.

When developing innovative solutions minimum viable products (MVPs) may not have expected results leading to constant changes in infrastructure, platform software and attached services. What if the MVP created does not fit target users need? In the first scenario, you already paid for all the resources you needed, so even if you shut down the solution, you already spent capex. However, in the second scenario (cloud), you paid for the service, so it's just opex. What if the solution had a great fit and people are using it a lot? In the first scenario, you have to follow an easier (but still hard) path to provision more resources, while in cloud computing, you can scale the solution resources whenever they are needed, not all the time.

Cloud computing also provides the ability to test new services and create new applications using best-of-breed services and technologies such as the IoT, cognitive computing, blockchain, and other services that are created and delivered on the internet. Those services are powerful tools to create new solutions and business models, providing a differentiated experience to clients whether there are 1, 10, 100, or 1 million people using the same service.