Cloud Computing vs Application Service Providers

One of the most common misunderstandings in just about any client I have is that Cloud Computing is simply a rebranding of Application Service Providers. At a very high level I suppose that it addresses a similar set of problems, however for the most part this is where the comparison ends.

For the people that were not playing around with Application Service Providers in 2000, it was simply where a company provided an application to clients to use, where the application and hardware to run the application was hosted by the company providing the service. So things like Email services using Microsoft Exchange, or SAP, or some large application was installed on a set of servers at the companies site, and the client paid rent to use this application.

The idea was great, clients got a running application, however they did not need to capitalise all of the hardware and procure the software, then simply signed an rental/lease agreement to use the service.

The biggest problem was the economics or the cost of the service. The application service providers set up a new set of servers and application for each client, and this would take some time to setup, and the cost was less than it would be for the client, however not that much cheaper.

There were two main issues:

- Barrier to entry: Clients could not trial the service, and it took a while to setup, you also needed to be relatively large to be a client. Small businesses would struggle to spend enough to interest the providers, as the initial setup cost was too much.

- Ongoing cost: The ongoing cost of the service was not that much cheaper than simply installing the software yourself.

Salesforce.com was the first big attempt at fixing both of these problems. From the start the design needed to be multi-tenanted (all clients on the one set of hardware and software). This simple concept took care of both the barrier to entry and the ongoing cost problem. The architecture of salesforce.com meant that although all the clients were on the same hardware and software, they did not know it. No individual client could see each other's data, even thought the data was stored on the same hardware.

Back in 2003, when I first started working with salesforce.com as a consultant and system integrator, it became quickly apparent that this new way of delivering applications or 'no software' solutions was vastly superior to application server provider based solutions.

Clients could trial the software in an instant. They could simply sign up to the service for a trial, and they could be using the application in minutes. So the barrier to entry issue had been removed. (Note, there was still a lot of data migration to do, however to trial the application was very fast). The ongoing cost was also a lot lower than the real costs.

This was the 'perfect' solution. An application that was available on-demand, and the ongoing cost was relatively cheap to an individual client compared to doing it themselves. The only concession on behalf of the client was that they could not customise the solution, they basically needed to take the customer relationship management solution as it was provided. While this was seen as an issue, in reality it was the right answer. Any client that was spending much time customising their CRM solution was not really focussing on providing a differentiated solution or (Unique Selling Proposition) for their clients. CRM was commoditising and clients needed to get on board.

This simple idea, in terms of providing a service on the Internet (in the Cloud) now comes in many forms (technology solutions, to human based solutions). However the core of the offer is providing you a simply (non customised) solution to deliver value to your business. It is still up to you to understand what you should be doing so that the simple cloud solution will actually deliver value, not just another faster way to spend money.