Cloudly with a Chance of Computing

The marketing team that originally went with “cloud computing” to describe various information services that are hosted outside of an organization’s walls may not have read The Clouds by Aristophanes.  In fact, they may have been inspired by the many Visio drawings of network engineers that had a little cloud to represent some wide area network.  Or maybe the originators of this appellation all live in a cloudy city where it rains all the time (like Seattle).  The truth is that it is difficult to know where these sorts of things get started.  But I was reading the lamentations of another writer about the return on investment of cloud computing (or should I say, lack of ROI), and his woes got me to thinking about clouds and whose computing they might actually benefit.

As a species, technology people are a suspicious bunch.  And many have control issues.  This may be the single largest cultural reason why organizations have such a hard time letting go of their core infrastructure or applications to an external vendor.

In spite of this, I don’t think any IT staff person would propose building their own search engine for the internet; everybody uses a “cloud” search engine like google or yahoo or bing to find stuff on the internet.  Even though the search engines actually keep track of what you are searching for and use that information to fine-tune their index (and potentially respond to subpoenas from people looking to sue or arrest you), I don’t think all that many IT people would go to senior management and say “I want $x billion in my budget to create a secure search engine for the internet for our organization.”  That’s because my hypothetical is silly.  Google is basically free (because it is paid for by advertisers), indexes an enormous amount of the internet, and is both relatively reliable (only one or two outages here and there) and relatively accurate in the results it returns.  Free is usually hard to beat, and when you throw in reliable with free, yep, time for a new project idea.

Now, a fair number of IT professionals are not thinking about internet search when they are considering a migration to cloud computing in their organizations.  I’d guess that core applications are on the list, like email, telephone services, document management, or other mission critical systems.  For those organizations that have gone through the pain of Microsoft Exchange 5.5, and the subsequent migrations to 2000, 2003, 2007, and maybe 2010, have experienced IT staff that can take a part and put back together an Exchange implementation, and experience little downtime today, cloud computing probably doesn’t make this better or less expensive.

Instead, cloud computing (like its older brother, the poisonous snake, the Application Service Provider (ASP)) is aimed at a different market segment.  For smaller organizations, who can’t afford a full time IT person and certainly aren’t going to pay for an Exchange specialist to be on staff, a cloud vendor is a reasonable alternative.  At $8 a month per mailbox, an organization of 10 users will pay a $1,000 per year to have their Exchange server hosted by a service provider like mailstreet.net – far less than the cost of an IT person and all of the licensing and equipment needed to host a server in house, never mind the backup, disaster recovery, and virus protection/anti-spyware services.

There are certainly downsides to placing your email with a cloud (what if the service provider goes out of business, what happens when you lose your internet connection, what if the service provider’s engineers keep reading your email), but I have a hint for IT people – most companies like cheaper whenever they can get away with it.  And in this case, spending $100,000 a year to have email in-house is hardly a good idea if you can do it for 1/100th of that cost by contract, unless of course your email (75% of which is spam and viruses) is so uber-important that you must have complete control over it.

The return on investment analysis will be different for more complex and proprietary systems.  While there may be plenty of cloud computing services offering you Microsoft Exchange, there are probably relatively few that will be able to offer hosting for the custom practice management systems for attorneys, or health records systems for physicians.  There are also more security and operational considerations for those sorts of systems – and a lower chance that a hosting provider will have the specialists on staff that you need to support that kind of system.  Notably, Lexis and Westlaw both provide what is essentially a hosted research service, and they are large enough to have teams of attorneys on staff to provide technical support to lawyers that use these services, but they appear to be the exception rather than the norm when it comes to other specialized systems.

So, in sum, cloud computing is aimed at providing services for organizations that can’t afford to host a system in-house, but can’t operate without access to the functionality of a particular application.  For organizations with existing IT staff and systems that work, I don’t see cloud computing easily supplanting either.  But then, google might release something that you can’t live without soon enough!