Monday, November 23, 2009

Cloud Computing: my thoughts

I love the Gartner Hype Cycle for Emerging Technologies. The 2009 version lists cloud computing at the 'Peak of Inflated Expectations.' Which means that, we are up for disappointment. But are we?

While several aspects of cloud computing are completely new, I would argue that it has been around for a while. When I first joined in 2002, it was an Application Service Provider (ASP), providing technology for companies to manage their transportation needs online, without a need for installing any software. It was soon acquired by Manhattan Associates which provided installed software. We worked for an year to enable our software to be packaged, and installed at the client's site. In 2007, we were back to where we starting, positioning our Software as a Service (SaaS) capabilities to target medium sized customers, who did not want to install software on-premises. Of course, this is very similar to Cloud Computing, as far as Business Software is concerned. I am sure MANH would be repositioning the software as a Cloud Computing Solution.

But back to the original question: are we in for a disappointment? I would argue that there are several factors working for and against cloud computing:


Sales people - yes you heard that right. Salespeople. Like I was. Why? Very simple - most companies haven't figured out a model for paying salespeople for selling Cloud Computing technology, especially as payments come a recurring installments rather than lumpsum payments. Ask any sales guy what he would want - a $100K payment in lumpsum, or four payments of $30K per year for the next four years? The 100K would be much preferred

IT Departments - IT Departments have much less work to do if the application is delivered over the cloud; in effect, you are outsourcing most of the IT maintainence work to the provider.In addition, IT Departments hate to lost control.

Economics - For major applications, like ERP, supply chain systems, I would argue that the economics do not vary much between buying and maintain the software vs. using it SaaS, especially for larger companies. Think of it; fundamentally what changes? The software is still being hosted on a server and database server. The server and DB server are still very efficiently used in the installed environment, because of the high needs of the company. So in one way or the other, things need to be paid for


Flexibility & Scalability - This is the single biggest driver for cloud computing. Several scenarios work very well for this, but perhaps no better than if you launch a new startup, you can just pay for usage. If things go well, your application is used more, your bills (and hopefully revenue) increase. If things do not go well, just end your service, pay your final bill, and you are done.

Cost - If the application usage is limited, and a lot of different clients can work off the same server/group of servers and use the computing resources more efficiently, the cost automatically goes down. Include the cost of installation, and cloud computing suddenly sounds like a much better option

Try before you buy - doesn't need much explanation

What applications are then prime for Cloud usage?

  • File Storage - example, Rackspace's Cloud Drive service
  • Platform as a Service - Microsoft's Azure and Joynet cloud service, amongst the multiple services in this arena
  • Operating Systems/Browsers for Netbooks, like Google Chrome and Jolicloud
  • Collaboration and Document sharing - Google Docs

So on and so forth.....

BTW I just realized, the Wiki overview of Cloud Computing Covers this and a lot more. So read it.

Another great resource explaining SaaS and PaaS is the Keynote address of salesforce CEO at Oracle Open World

1 comment:

  1. One of the challenges is the switching costs. Since large corporations have already invested lots of $$ in their existent infrastructure, switching could be costly...also i think there needs to be more thought on the economics of the subscription computin could have an impact on the PC industry as most of the storage and processing power will be transferred to the "cloud"..laptops might be significantly cheaper..and may be sale of netbooks could be a leading indicator of the success of the cloud computing partially...