The Web is sometimes referred to as the information highway where we can go online in search of relevant information we can use. Nothing about information retrieval turns out to be so easy.
In fact, the accuracy and relevancy of search results are unpredictable. Google and Bing never stop experimenting with algorithms and page layouts. A search engine’s main task is to deliver to each of us the most relevant and precise document based on each user-issued query.
Other factors influence the searcher experience, such as understandability, scope and reliability of information. When you see a series of search engine results, what you choose is based on what you need or what may be of interest. It’s important to understand the difference if you’re optimizing content for search or presenting information on a web page for your users. The most neglected aspect of content writing is findability and creating a stable, intuitive information architecture because taxonomies and word usage are not universally understood.
The Search for Relevance
The most ignored and unpredictable part of the information search experience is the obvious point that we are people and we are unique. A relevant search result for me will not necessarily be a relevant search result for you. I will ask a question differently from you. I may need a precise response and you may be browsing for something of interest. We expect an accurate response to our search question, but our search behaviors influence the results. Location, scope, language, interest, biases, interest, emotions and need can all factor into every search experience.
I recently moved from one home to another. One conversation held in the kitchen reminded me of how confusing communication is. My husband asked my daughter’s boyfriend if he had seen the “hand truck” anywhere. Here was the exchange.
Husband asks, “Have you seen the hand truck anywhere?”
Boyfriend responds, “I saw a dolly out in the driveway.”
Husband then says, “A dolly? You mean the hand truck.”
Boyfriend looks confused. “I call it a dolly.”
Husband, “I call it a hand truck.”
I interject with, “A dolly is what I played with when I was a little girl.”
If you have a website that sells that product, what do you call it? A hand truck, dolly or something else? What will your customers call it?
Prediction is a Mystery
Relevant results are supposed to be interesting, useful and easy to understand. We have all experienced false starts. When trust is broken, we develop new search behaviors such as ignoring Google’s paid top positions because there are no guarantees those top results deliver the information we need. They may be items of interest, but our primary task is meeting a precise need first.
One of the ways content writers try to engage us is by presenting interesting information that we can use first. When we are hooked, more documents are provided, or tasks appear such as calls to action. Page layouts change for devices, but a wise designer studies user behavior for clues on what information is of use the most often.
For example, when determining the order in which navigation links are presented in header global navigation, one common mistake is not placing what users need first in the lineup. Another common mistake is using vague terms that don’t describe the topic of a page or where a user will land if they click.
Predicting what users need requires time. Analytics help. So, does user testing. However, nothing is more valuable than truly knowing your specific customers’ behaviors.
The other day I was at the dentist for a routine teeth cleaning. My bubbly dental hygienist is always fun to listen to and on that day, she told me that she observed something unusual about her patients over a period of over a dozen years. She said that they don’t know it, but when she flosses their teeth, in most cases the tongue “points” to the next tooth to be flossed. She can’t explain it, but she marveled at how the tongue seemed to “know” where to guide her next.
I’m content mapping a series of websites and during this very painstaking, detail-oriented work, I’ve found specific information missing that people rely on for making decisions on where to go next, what information is timely, what content is relevant to whom, and even assistance with understanding topics and relevancy are missing.
What is provided in abundance is content that may meet a need and related content of interest, but none of it is targeted to specific user types. There is no help for information recall, which means that scanning may generate a hit but returning for something else that might also be relevant is a missed opportunity. We measure “time on page” but neglect to add assistance to increase that time or guidance to return for more. This could be something as simple as providing a way to add an article to a reading list or emailing content for later retrieval or sharing. Remember, “meet the need first, then interest.”
Another way to enhance the information retrieval user experience is to provide a way to show how many documents are relevant to your users. How many blog posts were made on a specific topic? How can they be sorted into sub-topics? How can you assign metadata to help define content and can you make it easy to prioritize by filters that matter to your users?
As much as we would like to believe we know how our users will respond to our websites, the truth is that their responses change based on what they already know and what they want to know, and both are always changing.
Ambient Findability, Book by Peter Morville
Understanding an enriched multidimensional user relevance model by analyzing query logs (http://onlinelibrary.wiley.com/doi/10.1002/asi.23868/full)