Bruce Byfield Took His Eyes Off the Prize

Mr. Byfield posted an article over at Datamation here. I must respectfully disagree with his analysis. Here’s why –

Look, even free costs money. Ask any father that has had a 4yo daughter bring home a stray cat. The fact is over the life of a product the support costs are the big dollar number not the buy. Having worked in the Enterprise Space I can tell you after getting your list down to potential candidates the next item on the check list is exit strategy! Not knowing how you will get out at the end its attendant costs can cost you many fold the entry price of the product. Ask anybody who used MS’s Enterprise Commerce Server, which is no longer available.

The allure of FOSS is not that you don’t spend any money on entry. Smart enterprises will get a support contract if they can for things like Red Hat, Sun Office, etc. Having that support pays dividends in saving in-house labor on searching for a solution. It helps mitigate the life cycle costs of the product. The real allure of FOSS is a reduction in lock in. Consider a product like Tivoli over say Nagios. Tivoli is an excellent product. But they have a game they play. The tool is broken up into salable components. It always seems that next component is in that next bundle. They don’t sell it ala carte. The database is proprietary, and transition costs to another product expensive. Nagios on the other hand is a use what you need philosophy. But your team will burn hours getting up to speed on Nagios. With Tivoli management can just write a check and have it deployed.

With the myth of ‘free’ dispensed with, still why use SaaS like services. Well in a nutshell ROI. The last company I worked for always had a quandary as to providing the necessary tools to everyone vs the attendant costs and support for those tools. The classic example is MS Project. In a perfect world everybody would get a copy. But why? In a large company less than 15% of the seats really need it. Not only that but tools like Project scream to be a shared platform choice. Project gains leverage when the project plan can be interactively edited by all concerned. Given those two factors using such a tool as a SaaS makes sense. You pay for the proper concurrent usage, share it online and when you are thru with it you discontinue use and the expense. The savings can be dramatic.

Finally there is the matter of ’scale’. Desktop applications are very different animals than something like Google or Twitter. You could not imagine something like Google as a desktop metaphor, the Google ToolBar notwithstanding. SaaS today has glommed onto tools like spreadsheets and Writers as it is a common metaphor with a large user base. But Saas will evolve and that evolution will look less and less like the desktop and more and more like Twitter, FaceBook and other interactive mediums. It is the only way one can scale and disperse the usage of such applications.

I might agree that for an Office Suite an Open Office on every desktop might make sense. But again, Open Office has the very nice trick of being a shareable tool. So why not load it on a in-house server and concentrate the admin costs to a single back end system. But the road is clear, having pushed applications out of the data center so people could get work done, we are now swinging back to the data center to help mitigate support costs. For common tools on every desktop leave them on virtual servers in-house. Those tools that are specialized, used infrequently and need to be shared. Rent, don’t buy.