The Forgotten Factory


Imagine for a moment that your company has built a factory and filled it with machinery. Or perhaps your company has built a large, new office building with a sizeable parking lot and a good cafeteria. These additions represent considerable investment. Now, imagine that they have forgotten where they built the factory and where the office is located. Beyond imagination, you say. No company would be that incompetent.

Next, imagine that over the last 10 years, you and your colleagues have been writing reports on competitors, market prospects, new products ideas, and business strategies. These reports have been filed away on a shared drive or an intranet. But now, because your search engine is not fit for its purpose, you cannot find the information that your company has invested in creating. Would you say that it is not possible to imagine and that your company would not be that incompetent?

I now want you to put this copy of EContent aside (or open a new browser window) and go to, where you will find a ticker display of the growth rate of digital information developed for EMC by International Data Corp. The rate at which the number escalates in front of your eyes makes the U.S. National Debt ticker look slow by comparison. Whether the number is correct to the nth significant figure is irrelevant. In your company, you are creating content at an alarming rate, and yet you can’t find information with any degree of confidence.

Late last year, Jane McConnell published her annual "Global Intranet Trends" report (, which indicated that companies are beginning to put more support and thought into intranets. The notable exception was search. Sixty percent of companies regarded search as business-critical, and yet only 10% of companies reported that their search was anything more than satisfactory. To play another "imagine" game: How long would you stay in business if your customers only gave you a 10% vote of confidence?

Time for more bad news. All the indications are that 2009 will be the year that enterprise applications of MOSS07 will start to ramp up. Whether this is a good strategy is not the question. What MOSS07 will do is make it even easier to amass Microsoft Office files. According to Microsoft, there are 100 million client licenses for the product. So figure out the amount of content you create a year and multiply it by 100 million—that ticker number will speed up exponentially. I think Microsoft did the same calculation, figured out that the search application in MOSS07 wouldn’t be able to keep up at enterprise level, and hastily rushed out to buy FAST Search & Transfer.

Over the last couple of years, I’ve developed various approaches to developing a business case for search. But reading McConnell’s report, I realized I’d missed the obvious reason. One of the major reasons for introducing search is to compensate for the inadequacies of the information architecture (IA) of the intranet. Employees cannot find anything using the IA, so we’ll give them a search engine. Full stop.

What I mean by full stop is that all that will happen is a search application is installed, and then, at best, someone comes by from time to time to dust the servers. Time and time again, I find that a failure of search is an organizational rather than a technical issue. Even a Google search appliance needs more than an occasional dusting, although Google would like you to think otherwise.

As you may have gathered by now, I get fairly passionate about search. I’ve been involved with search technology since about 1974, and I still find that the fundamental issues are not well-understood by vendors, not to mention users. Finding relevant enterprise search-related articles and other resources is certainly a good start, but when defending a patent or responding to a compliance request, "find all" is everything. How many times in the last couple of months have you searched for information and then wondered to yourself, "Don’t we have more than this?" The answer is that you almost certainly do. If you want to find the factory or the office you constructed, then keep reading my Eureka columns and I’ll see what I can do to help you. The solution does not start with search but with information management. But that is for my next column.