Enterprise search is a strategy, not a technology. It is a strategy for ensuring that employees and customers can find information that an organization has created.

As I was choosing the topics for the recently published second edition of my book "Enterprise Search," I decided to include a new chapter on website search. Chapter 9 of the recently published 4th edition of "Information Architecture" includes the definitive advice on website search. But my objective in including website search in the new edition was because it is usually seen as the responsibility of the website owners and left out of the enterprise search strategy. This means that the skills of the enterprise search team may not be readily available to the website team, to whom search often appears to be an afterthought.

Poor Results from Law Firm Websites

A current hot topic in international law is the future of US/EU traffic in personal data. This follows the European Court of Justice ruling on the legal validity of the FTC Safe Harbor protocols. This ruling carries with it significant implications, so chances are high that senior executives are looking for some advice on the actions they should take. So I went to the top ten law firms as listed in Wikipedia and used "data privacy" as a query on their websites.

The poor quality of the search experience surprised me — I found five areas where these sites failed to deliver even close to current good practice. 

1. Several responded to the query by displaying lawyers with relevant experience, without any introduction to the practice areas of the firm. Existing clients already have a client contact to introduce them to lawyers if they need help. For those who are not yet clients, presenting a list of lawyers without the surrounding context is not going to be of immediate assistance.

2. The results lacked any indication of a date on the items returned and the option to sort by date. The ECJ opinion has undone a great deal of good practice. I would have expected the largest international firms to be promoting the most recent advice they can offer. 

3. The results also included filters that were ineffective in reducing very substantial number of results to a reasonable level for review.

4. The quality of document titles was poor and variable — for example, some of the results on the law firm Freshfields website do not have a title. 

5. Finally comes the issue of plurals. Law firm Linklaters website offers 726 results for "contract" and 2865 for "contracts."

Lessons to Be Learned

You would expect international law firms to be paying very careful attention to information quality, as it is the basis of the value they provide their clients. Would you wish to do business with a firm — any firm — that erects barriers to finding information on its own website? There seems to be no clear view of search's value as an element of website information architecture. While often well constructed in a "browse" context, no one from the communications department seems to have considered how important search on the website can be. They fail to see its usefulness in gaining new customers who want to see the quality of relevant briefing papers before picking up the phone to a lawyer.

I suspect that one of the reasons for this is that the website has been "designed." Design firms have little understanding of the role of search and in the course of user requirements research or user testing, will see a user's request for search as a failure of the site architecture. Read "Information Architecture," and it will place search in context of site design.

The advice I give to marketing and communications departments is to take queries at both the top and bottom end of the use frequency logs, and run the queries together with some line-of-business managers (senior partners in law firms). Then discuss whether the website search is fit for purpose in delivering an exemplary search experience to potential clients.

Back to Enterprise Search

When I see results like these on an organization’s website search, it makes me wonder about the effectiveness of its internal search applications. There seem to be three options: If the internal applications are substantially better, why was so little attention paid to website search? If they are no better than the website search, this points to a significant lack of commitment to managing information as an asset. And then there's the third option — internal search could be even worse!

Which option has your organization adopted?