Search: Give It More Thought

//Search: Give It More Thought

In many of my information architecture projects, planning and designing the search results page isn’t given much thought. While search engine optimization is a big thing for users to find the site, searching within a site and displaying effective search results are seen as a “plug and play” thing. You simply “turn on” the search feature, the site gives you a ready made page, and this is what you use.

Unfortunately, with so much content on a website, a simple search results page normally isn’t good enough. The search results themselves need to be optimized in some kind of way. I can give you one example: search engines normally do a full text search, meaning that any term you type into a search box will return results with that term anywhere in the document. If you typed, “Tiffanys” into a search engine, you would get results about Tiffany’s (the jewelry store), Breakfast at Tiffany’s (the movie), and about the million Tiffany’s in the world. But what are you really looking for? Is it that diamond ring or more information about Audrey Hepburn? Do you want references from this year or do you want archives? Do you want images or text references? How is ambiguity resolved?

Search results need to be given more thought.

Information Seeking

Previously I posted to articles on this site about Information Seeking Fundamentals and Applying Information Seeking Principles. In the first article, I wrote about berrypicking: people will pick their way through different information sources to find what they need. They’ll follow a trail (such as footnotes or hyperlinks) and look at related topics. But they’ll also avoid information they don’t want to see, don’t find useful, or comes from a site or person they don’t like. In the second article, I touched on information overload. Sometimes people feel overwhelmed by the amount of content available and they simply can’t sort through it all.

To conclude these two articles, I noted:

As corporations produce more information, vie for more of their users’ head space and ask people to do something with the information provided, it will become more important to reflect the users’ mental models and information seeking processes in our designs. If we want to make information actionable, it needs to be findable and comprehensible.

Nielsen’s Article on Search Skills

In 2011, Jakob Nielsen posted an Alertbox on search skills called Incompetent Research Skills Curb Users’ Problem Solving. While I may go on about information seeking skills, his research shows that many people don’t have strong searching skills and use simple searches to find information.

Nielsen identifies that many search boxes are undermining the user’s ability to find information to problem solve. Users continual use the search box and don’t change strategies.

Still, the rough estimate from our available data is obvious: users change search strategy only 1% of the time; 99% of the time they plod along a single unwavering path. Whether the true number is 2% or 0.5%, the big-picture conclusion is the same: users have extraordinarily inadequate research skills when it comes to solving problems on the Web.

As part of these “extraordinarily inadequate research skills,” users also rarely access advanced search and if they do, they use it incorrectly. In my experience, advanced search can be quite complicated. Everyday people don’t keep Boolean logic in their heads and don’t memorize search commands, so it’s no wonder why advanced search is used incorrectly.

Federated Search

Today’s search engines, such as Google or Yahoo or Bing!, really give us a federated search, of sorts. Federated searching can be useful to give us an idea of the information available. But these searches tell us little about the quality of the information returned. They also make refinement difficult since the different sources use different metadata.

Corporate Searching

To circle back around to implementing search results within your company, it’s important to take all these things into account. Recognize the different information seeking behaviours and motivations; realize that people have little to no training or ability to perform complex searches or to change their search tactics; recognize that advanced search isn’t used. has an article on Enterprise Search and Why We Can’t Just Google. In this article, they emphasize the same points as this post. Full text search doesn’t work and we need to provide better search tools within the enterprise.

…we need to change our perspective on what we expect from enterprise search based on what we’re willing to do to make it work. This means taking a closer look into redesigning the overall experience to move away from an emphasis on full-text indexing and toward ways that not only provide direct access to the answer, but also promote discovery, exploration and raise awareness. Enterprise search should in fact be more relevant inside the organization primarily because we have greater control over both the inputs and outputs required to make it work.

As part of helping people perform better searches, we can use metadata and taxonomy refinement. The article also gives six steps to helping improve enterprise search.


This article is meant to raise awareness of the challenge of search results within the corporation (or enterprise). A search tool that is quickly implemented without any testing or refinement will most likely not serve the needs of the user. We need to ask: what search skills do our employees have? How can we match these skills within search? How can we teach better search skills? What can we do to search to improve result rankings and give better results?

By | 2018-01-16T16:12:24+00:00 July 3rd, 2012|Categories: Information Architecture|2 Comments


  1. Nicholas Herold July 4, 2012 at 10:27 am

    Excellent article. Thanks, Theresa, for a really thoughtful piece about the human end of searching. There are now many “curation” tools available, each with their own strengths and weaknesses, but what really disturbs me is the reliance on automated systems to perform discovery tasks.

    As you note, we can’t just Google anymore, but not only because of the reasons you mention. There are two important characteristics of discovery results:: 1) how is relevance determined? and 2) what are you missing? Using federated results still relies on the “federation” of users or raters understands the relative importance of what you are looking for.

    If your federation is made up of experts and non-experts, relevance gets diluted. And from a competitive perspective, knowing things that your competitors don’t know, or finding them out earlier than they do is an advantage. If you’re using the same discovery process, you’re pretty much limited to the same results, with the same list structure as they are.

    A successful discovery system should make you aware of all possible results, but in a way that allows simple and speedy weeding out. It should also allow you to be surprised by results that you determine as relevant, but that you weren’t specifically searching for.

    Oh, and did I mention the results should be presented in real time?

  2. Theresa Putkey July 12, 2012 at 9:19 am

    Thanks for your comments, Nicholas. There are many many aspects to the search experience, I completely agree with you. Everything you said has the makings of (at least) a few more articles!

Comments are closed.