The StoneHenge blog

When choosing a web content management system, what should you look for? Too often I see this question devolve into a debate over "open source" vs. "commercial," ".NET" vs. "Java," or worse yet, "SharePoint" vs."everything else."

I say this not because I despise SharePoint -- I do, but that's a whole 'nother subject. No, the real problem is that this debate puts the cart before the horse. It's like arguing the relative merits of driving vs. flying: First, where do you want to go? 

In my experience, the key to a successful CMS implementation isn't in the code -- be it roll-your-own, open source or commercial. It's in the information architecture, content modeling, planning & analysis. Key questions:

1. Size: How many pages do you have? How are they related? How often will you update content? How much media, and how much of it is video?
2. Workflow: How many authors will you have? How many approvers? How many approval steps will a content item have to pass before publishing? Do you require an audit trail and unlimited versioning?
3. Social networking: Do you want to use your website as a two-way conversation with your customers? Will you have a blog? Comments? Community forum? Wiki?
4. Back-end integration: How many databases will you tie into for things like ordering products? How many third-party apps will you tie into, like PeopleSoft for HR job listings?
5. Personalization: Will your customers log into the site? Will you keep customer data profiles? Will you run permission-based email marketing?

Answering these questions is time-consuming, often difficult, and can suck in a small swarm of stakeholders of the company, but is vital for a successful CMS. Without it, you could spend months implementing a system that no one will use because while it does many cool things, it doesn't do the thing you need it to do.

Here's the right way. Earlier this year I completed a content model that ran well over 200 pages. After it was finished, then the client selected a CMS that could meet its requirements. (They chose Sitecore, by the way, but mostly because they were an ASP.NET shop.) The implementation went well, but even more important, today they are using their system. Happily ever after.