Some people assume that if a concept is old, it's no longer relevant. Since we looked at the continued relevancy of the quaint old fashioned concept of information publishing on intranets in last month's post, I thought we'd take on a few more of those dusty relics of days gone by today: metadata and intranet search. The concepts are interrelated in that the success of communicating content on a one to many basis -- or an n to m basis, as you may have a small body of publishers or content creators putting out information to the company -- relies on findability.
In order to consume information, your employees have to find it. You can help them do this in many ways: from the structure of your portals and intranet sites, through the creation of indexes and site maps, to the thorough application of metadata to the content items. To improve it's "ambient findability" content should be well structured itself, and provided with plenty of appropriate metadata. (Check out the book "Ambient Findability" by Peter Mooreville).
Unfortunately this seems to have become an old fashioned concept in some quarters. The idea we're hearing from the social collaboration world is that structured metadata schemas and the building of taxonomies are "old school." I would argue that this is simply not so, and the free-for-all that is an enterprise folksonomy is not going to be successful in every scenario (or many of them).
The Need for Good Metadata
Once an intranet or intranet site or sites reaches a certain size it will become increasing difficult for people to find the pertinent content they require at any given time to complete any given task. At this point we should see a good collaboration between UX design specialists and information management professionals.
It is a popular belief (based on research?) that people fall into two overarching categories: those who like to browse a navigation structure to find information and those who just like to use search tools / interfaces. Some even suggest this is generational, in that those of us who grew up with early versions of Windows and Mac OS (or even Unix) are used to browsing folder structures, while the post-Google generation are much more used to a simple search box. I am not sure that it's so clear cut along those generational lines, but I do think it's a case of the right tool, or the right interface, for the right job.
For the information publishers, being able to browse a structure might help them put the information not just in the right "place" but in the right context, particularly if the CMS can automatically append metadata based on which container the content item has been placed. As to whether or not a "simple search box" cuts the mustard on an intranet, let's return to that later.
There are scenarios where a folksonomy should work just fine. And by this, I mean a true one, where the user gets to pick any value they like as a "tag." However, these tend to work better on the public internet due to scale and critical mass. On an intranet they may work at the smaller end of the sale, with wikis and other forms of user generated content including within workspaces or microblogging elements of a social collaboration platform.
Intranet Search versus Enterprise Search
If intranet publishing is not dead as a concept within the modern digital workplace, then does the concept of a simple search facility for published content also remain valid ? Last month I used two platforms --ThoughtFarmer and Jostle -- as examples of modern intranet platforms that retain functionality for information publishing as well as social collaboration. They also have built in search functionality, that uses metadata to help users find the pages they want, or the blog posts, status updates or people they might be looking for.
On all but the smallest intranets, the simple search engine that only provides access to the Web Content Management System and its content may be a thing of the past (or fast approaching it, if we are to believe the search engine vendors). The vendors may be right in some respects, but if your organizational strategy is based on each main line of business application having its own built-in search functionality, that strategy may be extended to the "publishing intranets" WCM platform and to any separate social collaboration system. You can attempt to pull them all together using federated search and providing a single interface or search portal to improve the user experience.
Or you might go for a single search engine to rule them all. You can use a SharePoint, a Google Search Appliance or a good mid-range tool like Coveo to index multiple content sources to provide a single back end and index, as well as single interface. Returning to metadata, many modern search engines can crack open common types of content object, parse the contents and automatically (or is that auto-magically?) append metadata values -- but be warned your mileage may vary with the efficacy of such technology depending on your organizational complexity, etc.
Concentrating on metadata when publishing content to an intranet is no more antiquated than the concept of web publishing with a digital workplace is. Enterprise Information Management remains a broad and complex domain, and while some organizations may succeed by ditching everything in favor of a social collaboration platform and self-tagging folksonomies, it probably won't work out well for the majority of us!