Do you think that cloud computing will eventually replace conventional software?

Yes--for 80% of software (The Pareto "80/20" applies here, like in many places) by 2020.

Cloud computing offers too many advantages: 1) elimination of capital budgeting; 2) flexibility to turn usage up, down -- or off -- on-demand; 3) one-stop access to continuous R&D, testing, maintenance, upgrade, disaster recovery, security management, etc.; 4) lower cost economy-of-scale advantages, and more.

80% of people will be able to anything they need on a public, private or community cloud. Just as as 80% of the use of your computer today is conducted through a browser. Only those 20% of applications that a industry-unique, high risk mission critical, etc. will use legacy architectures.

An easier way to think of this is to look at cloud computing as a utility: how many business own their own power stations (vs. pay for electricity on-demand)?

From a marketing perspective, the on-demand rental of tube sock software applications, billed via the usage, seat, subscription and click thru process will continue to displace the conventional 'packaged' and ‘shrink wrap’ software license pricing and conventional life cycle software models.

In fact, the market shift to the cloud computing pricing and life cycle model is accelerating and it seems inevitable for most new software. Who in the future will prepay a $600K licensing fee to set up a SAP pilot project? Emerging nations? Fortune 50? Medical/government institutions? Maybe they will, but it is surely a shrinking market.

The economics and conveniences of cloud computing commercial fee models are compelling, versus the conventional prepayment license fee, the per 'customer' customization fees, the test/debug fees, the pilot project support costs and the additional fees for maintenance and new release upgrades.

Does the modified cloud computing market model replace conventional software or just make it impractical to sell new starts, using the conventional licensing model in the cloud computing modified market?

The new Service Oriented Architecture (SOA), modeling processes and languages that are driving Cloud Computing applications and appliances are already displacing some portions of the conventional software prepay environments.

While fortune fifty, government and medical institutions might continue to pay the conventional fee structures, for certain types of customized programming over the next several years, they will (are already) shifting large amounts of their IT to the tube sock cloud computing model.

Will these large entities be able to leverage the replacement programming cost from the software vendors and pay them later with Cloud Computing charges, under long term contract?

At the profit levels of Oracle, Microsoft, IBM and SAP, I would think account control needs in these massive accounts will push the suppliers to provide the custom work, without the front end pre-payment of conventional software fees. When that happens, conventional software licensing and fees will formally pass into antiquity.

Not completely, but mostly. I agree with Jim - 80% is a pretty good guesstimate.

For packaged applications the vast majority will move to the Cloud - the economics are simply too powerful for organizations to continue to justify licensing and managing these applications on premise. As Jim suggested above, do companies manage their own power generation? Do they run their own telephone networks? Packaged applications are going the same direction - they will be utilities delivered online via the cloud.

For custom applications, the percentage will be lower (although still significant - particularly within Private Clouds). Since these applications are by definition unique, the economic justification to maintain them in-house is more significant.

I'm with Jim on this I see adoption as a two staged process. First we'll get to pareto optimality (80/20) where we'll hover for awhile. That will take us a few years still.

Once we're there we'll see a trend towards zero kick in. As less and less software is delivered on-premise, the cost of doing so (and the obsolescence of that software) will force people to move more to the cloud. An analogy - while it's true that there are still mainframe computers around the place still in service, the number is greatly trending downwards and the move to cloud software is analogous with this...

My 2 cents

I have been writing about the benefits of migrating to the Cloud in previous articles but it is also important to highlight in which circumstances the Cloud Computing route may not be the appropriate one.

Here is a list in no specific order of ten good reasons why Public Cloud Computing may not be a good fit for your company:
http://www.getapp.com/blog/reasons-no-cloud-computing/

The short answer is Yes. Think about how much you rely on your internet connection right now. No internet = no email, no online research and no file transfers just to name a few. One of the biggest drawbacks I've read about cloud computing in the overreliance on your internet connection - well you're relying on it right now for essential business activities. And as wireless services become faster it is another reason that cloud computing will eventually replace desktop software.

Look at all of the advantages of cloud computing - (1) no updating software, (2) reduced IT costs, (3) anytime anywhere access, (4) easier scalability, (5) better collaboration. My guess is that within the next 5 to 10 years a majority of all computing will be in the cloud and it is just a matter of time before substantially all computing will be done in the cloud.

Yes, when critical mass is there and adoption takes place. Traditional IT will still exist but at the edges of the cloud.

No, not 100%.

There are plenty of situations that call for on-demand, web-enabled software, and these will continue to rise. However, there are plenty of circumstances which call for locally hosted software, and I don't think this will ever really go away until some completely new paradigm comes along.

Think of all the things that were supposed to be made obsolete, and have yet to.

Also consider this: I'll bet that there are very few cloud providers that obtain 100% of their own backoffice solutions from the cloud.

-ASB: http://about.me/Andrew.S.Baker

Jim is partially correct that companies dont produce their own electricity, however if they have not done their risk assessments then they leave themselves open to major issues. Hence companies may produce their own electricity if the grid is down via generators.

The cloud is just another option that has to be considered in risk mitigation. We have seen the good and the bad over the past few months with the natural disasters and the effect it has on the infrastructure. Some of our Christchurch (NZ) based customers were rewarded with going to the cloud and their branches were still online. But then we had issues in Queensland with the communications infrastructure being down and if they were on a On-Premise install they probably wouldn't have been offline for over a month.

So its something that needs considering.