Browse Strategies and Site Directory
Posted on April 18, 2008
by Joel Oleson
Leave a Comment
Now that I’m not at MS I can be a bit more candid about things I like and don’t like, I hope some of that came through in my previous blogs anyway, but want to let you know I have been honest, but maybe not brutally honest unless you were in my Good Bad Ugly sessions at TechReady5 and TechReady6 (internal technical readiness conferences).
The site directory out of the box is cute. It fits the really small environments out of the box, but for even some medium environments a company needs to really think about a browse strategy for getting people to their sites.
What doesn’t work with Site Directory
I. All sites are added automatically generating at least a few problems.
1. Sites that are empty are added
2. Users that add their own descriptions can put some sites over the top in terms of scope. Data may not be relevant.
3. If you click on a site you more than half the time will not have access. How many times does it take to make you not want to use the site directory? Sites are not security trimmed.
4. Sites may no longer exist (some features in MOSS 2007 can help you take care of this one at least.)
What you need…
1. Sites that are useful and interesting… not old, not crusty, not empty, not shallow…
2. Sites that are security trimmed or are open
3. Categorizations need to be useful and easy to navigate and not empty nodes (not too deep)
MSWeb is one example of an interesting evolution of a site directory and browse strategy for the Intranet.
1. First before it was even SharePoint there was a list of sites that was very groomed and managed. People were able to find the sites that were approved, but you couldn’t find data that wasn’t exposed. Even search was only of the known universe.
Result… a site directory of the known sites that was easy to use, but was not automatic and was managed.
2. Then there was SharePoint 2003, I was anxious to see a site directory of all the sites at Microsoft. An all inclusive directory of all sites allowing it to expose all those hidden gems of data across the company. Search as well, was looking at true enterprise search or search of data across the company in all web based repositories including some limited file shares and public folders. A site creation UI was designed to dump all sites into the site directory and search would be based on a dump of all of these sites and the site directory.
Result: A huge site directory that would make you proud, except in SPS 2003 we had all sorts of empty sites, deleted sites, and irrelevant sites. Really you couldn’t find anything you were looking for, but the categories looked very impressive.
3. The next generation site directory would be scrubbed the hundreds of thousands of sites would be seen as irrelevant for the average person. If they needed something out of an obscure site, then they would either know the URL or they could search for it. This new directory would go back to the roots and start with the known universe. If you’re looking for the SharePoint Team, there’s one public facing SharePoint site for that team and so on. This more static, but managed list would be the stategy. No more would it be a full list of all sites. It would be open and official sites… not really a lot of collab sites, but more portal or WCM type sites.
Result: People started returning to using the site directory, but search really took over as the standard for finding data. Browse took a back seat. Browse still isn’t very big at MS, but there is some promise that these issues of manually managing such a list could provide promise for the future.
Like it? Why not share it?