Mircea hit the nail on the head here. Unless your company is offering something super exciting and revolutionary (something to warrant going viral), then you're going to have to plan it. Luckily it doesn't appear to be that hard to make a video go viral.
For example, the liquid mountaineering video: http://www.youtube.com/watch?v=Oe3St1GgoHQ - I think it was for shoes, don't remember, but it was some guys who apparently could run on water, when in reality you can't. It looked realistic and went super viral.
Another is the shaving helmet: http://www.youtube.com/watch?v=5bgRszdUdhQ And of course the most known and successful (thus far) viral campaign for a company, the Old Spice commercials. Come up with something that defies the odds, is cool and makes people say, "wow, that's so damn awesome."
Learn from other peoples viral successes and adapt to suit your needs.
I think it's realistic for it to be a component of the plan, but it's madness for it to be a critical component.
For me, the business should stand on it's own, and have a plan for success based upon organic growth, otherwise you just bought a lottery ticket!
By all means have a plan to try and encourage it going viral by making it easy to refer, send invites and the like, and to treat that as part of the process just like traditional marketing or maintaining the blog. But at the end of the day whether it actually works and actually goes viral depends on whether your business engages enough people to get critical mass.
You kinda have to plan it to happen. If we're talking about a business, not that many people go crazy about the launch of a new company otherwise. Or at least word doesn't spread that fast without one.
The question is how essential is it, how much does it cost, and your gut feeling if it will work.
There are 2 ways to become viral. Both have to be pre-planned
Plan 1) Having something fun and sharable, and unique that you plan to push (often with the help of an agency) via a large number of sites and mediums in the hope that it will generate enough buzz to create a viral effect. An example of this is the Old Spice viral campaign about a year ago.
Plan 2) Your content or product has to be social and sharable in nature. For example, it might be something people do socially off-line like having meals together or attending events.
The term 'going viral' actually has a specific mathematical meaning. We use it loosely to mean 'a bajillion people saw it', but it's what happens when a product has a sufficiently high viral coefficient. Making a product, webpage, or video 'go viral' means making something people pass on to their friends frequently enough that the content propagates through the network. If you have 1,000,000 views because you sent something to 10 people and they passed it on n times, you've gone viral. If you have 1,000,000 views because you bought a 100,000,000 email spam campaign and got 1% of the people to click the link, you have not gone viral.
There's uncertainty any time you try to 'make something go viral', but if you use the excellent monitoring and analytics tools we have at our disposal to measure the different parts of the viral coefficient you can optimize your product for 'virality.'
Here's an example.
You have an application that let's people post content to their Facebook walls from several places within your site. Your hope is that this feature will increase the number of people that your user exposes your product to, and as a result increase new users. You can include a link back to you product in the FB wall posting that includes a token, and track how many times each user uses the 'post to FB' feature, how many inbound visits each one generates, and how many of those inbound visitors convert to users. Google analytics and split testing packages will let you monitor how far through the process you get, and you can work to improve whichever step is failing to raise your viral coefficient.
In short, yes, you can and should plan on going viral, but you should plan on it taking a lot of work, not luck.