Search Blog:

To infinite search, and beyond

We're progressing down the Road to Social Knowledge Networks today, and cooling our heels at the next stop, infinite search.

This is often the next stage in an organization's content management strategy. Organizations turn to search as a way to address the No. 1 complaint they hear from workers regarding the shared network drive (or junk drawer): "I can't find anything."

The problem with search is that it fails to address the fundamental problems of poor quality content and basic disorganization. In fact, it's a lot like covering a plate of spaghetti with tomato sauce. It might look better, but the information is still a mess.

You are simply changing the place where you look for information, and move from sifting through the content on the shared drive, to sifting through the search results.

Why does search often fail to live up to its over-hyped promise? Due to one simple but often overlooked fact. Relevance is not the same as quality.

The search engine can rank and sort search results by relevance, but it cannot know anything about the quality of the document. If you want workers to find high-quality documents quickly, the search engine isn't your best option.

Here's an example. Say the shared drive contains a document drafted by marketing. Three months later, R&D revises the document and saves it under a new filename. Sales updates this version four months later, but makes several major errors and saves it under yet another filename.

The search engine will likely return all three documents when and if you do an appropriate search. They will all be relevant, because they contain the same content for the most part. But how does the worker understand that the R&D document is high quality, and the sales version has serious errors?

Just by the numbers, if 80 percent of the hard drive is outdated, irrelevant data, the search engine will yield 80 percent outdated, irrelevant data. You need a process and workflow that will do several things search engines don't do:
  • Delete outdated content and eliminate erroneous information.
  • Inject a quality metric into the system.
  • Encourage the sourcing of high-quality content.
Without this, you do not *really* help people find the high-quality content quickly. Rather, you help them find content of dubious quality quickly. We'll be discussing how to solve these issues in a later post as we move down the Road to Social Knowledge Networks.

Key take away: Relevance is not the same as quality. Plain search just lets you find documents of dubious quality more easily.

No comments:


Related Posts Plugin for WordPress, Blogger...