At Search Engine Strategies in San Jose this August, I had the good fortune of being able to share a panel with Nico Brooks from Atlas. I’ve always admired Nico as one of the more strategic thinkers in this space and feel a kindredship, both through our love of research and curiosity about user behavior. Nico brought up what he called a “Big Idea.” The concept was fascinating and I followed up with Nico after to pick his brain a little bit more and to do some further research into the topic. I still hope to work with Nico on a more thorough exploration of this topic, but I wouldn’t mind covering it at a high level in today’s column.
When it comes to how humans navigate a physical space, there are three types of knowledge we acquire along the way (for more, check out the academic paper that Nico brought to my attention). The first of these is landscape knowledge. We look for those reference points in the landscape that stand out from the surroundings and are quite often found at forks in the path. These landmarks sit at potential junctures in our route and we can draw our bearings from them. We turn left at the big yellow house with the white veranda in front. We hang a right at the McDonald’s on the corner. Landmarks help us get our bearings in unfamiliar space, and we use them as navigational points of reference. The first thing we do when we enter unfamiliar territory is look for landmarks.
The second level of knowledge is route knowledge. As we begin to navigate from landmark to landmark, we begin to memorize the routes we take. How often have you been in an unfamiliar city but you’ve memorized the route to get from point a to point b? Undoubtedly there are shorter paths to get there, but you stick to the route you’ve become familiar with. We memorize the steps we need to take to get to our destination. We know if we stay on this road until we see the Denny’s on the corner, then all we need to do is turn right, follow another road past the Exxon station and eventually, in this manner, we’ll get to where we want to go. Nico uses the example of taking his daughter for a walk in the stroller. The end destination was a playground not far from their house. But Nico, thinking he could work in a bit of a longer walk, took the indirect route that took them past several interesting points on the way. Eventually they went in almost a complete circle and ended up at the playground that was a few steps from the Brooks residence. The strategy worked well until Nico’s daughter started to acquire both landmark and route knowledge. Eventually, enroute, she recognized a landmark that she could see from her home and realized that Nico was trying to pull the wool over her eyes. The jig was up.
The third level of knowledge is survey knowledge. As we become more familiar with our online environment, we begin to create an abstract representation of both the routes and the landmarks. We use these to form a cognitive map. This cognitive map allows us to calculate the best route from several options. We have a rough idea of where roads go and how the grid of a city might be laid out. We become more adventurous in our navigation, knowing that even if we go astray, in our minds we have a relatively reliable mental representation of the landscape and sooner or later will get to either a recognizable route or landmark.
Is there a virtual translation?
Nico and I had a chance to chat about how this may translate to the online world. The difference is, of course, our online world doesn’t have physical dimensions. The idea of “space” has no place on the Internet. The distance from my desktop to a website for the realtor’s office I can see across the street from my office is no different than the distance from my desktop to a website for a hotel in Beijing, or a lawyer in Kraków, Poland. So, if we have no dimensions to interpret, how do we navigate online?
I think the first two levels of knowledge still apply. We still identify landmarks, and we still memorize routes. It’s at the third level, survey knowledge, where the analogy doesn’t translate particularly well.
Consider this. As we navigate unfamiliar territory online, the first thing we look for is landmarks. And often, these landmarks serve as navigational hubs in the online space. Of course, our favorite landmarks tend to be our preferred engine. Google, Yahoo, and other search engines stand as virtual megaliths in the online landscape. These landmarks are ingrained in us as the jumping off point for almost all our online navigation. And they act as the ultimate navigational hub. From here we can jump practically anywhere on the web. And we use these super-landmarks to find other landmarks.
If we’re venturing into unfamiliar territory, the first place we go is our favorite search engine. We launch a query that we believe will give us the best odds for identifying navigational landmarks in this new online landscape. Then, we scour the results page to see which landmarks our search engine has brought back for us. In some cases they could be vertical directories or search engines. It could be a well-known vendor in the space. It could be a recognized online publication. Our landmarks will vary somewhat depending on our intent, but the common characteristic is that we want our landmarks to be authoritative, authentic, recognizable, and, more often than not, a comprehensive navigational hub. It’s the last of these characteristics, comprehensiveness, which is the goal of blended search results currently being seen on Google (Universal search) and Ask (3D Search).
The search shortcut
Even if we’re treading down a familiar path, we’ll often use our search engine as a navigational shortcut to get there. Navigational searches are becoming a more and more common user behavior. If we need to navigate to a particular page deep within a website, most often it’s a lot easier to use a search engine to get there. We become more specific with our query, we scan the results for the URL we’re looking for, and we hope the link provided will take us to the right page within the website. We don’t have to worry about misspelling the URL, we don’t have to fight with an unfamiliar navigation scheme, and we can qualify our decision before we click by looking at the title and the description provided by the search engine to determine if this is actually the page we want to go to. In our recent B to B survey, we asked respondents how they would navigate online when they were beginning research into a particular business purchase. Slightly over half indicated the first place they turn would be a major search engine, with the lion’s share of this traffic going to Google. With all the rest of the answers, we then reframed the question and asked them how they would navigate to their other destinations, whether it be a vendor’s website, a directory they were familiar with, a vertical search engine, or other site. Once the question was reframed, it turned out that a full 25% of these people would actually navigate to their site of choice through a major search engine.
So, in almost all instances, we navigate online spaces by identifying and using landmarks. We don’t use them to get our physical bearings in this virtual environment, like we do in the real world. Bearings have no meaning if there is no physical space. We use these landmarks to get closer to the information we’re looking for. This brings us to another key difference between an online and off-line landscape, informational proximity rather than physical proximity, but I’ll get to this in a minute.
Remembering online routes
In thinking about this, I realized that we also acquire route knowledge online. It may not be done as consistently as it’s done in the off-line world, but often I get back to a site I visited previously by retracing the steps I took to get there. For example, I may not remember the URL of the site I visited or the page I was on, but I often remember the query I launched to get there, the approximate place on the search results page where I clicked, and enough other details to allow me to start with my favorite landmark, my search engine, and relaunch the search and follow my virtual trail of breadcrumbs to get back to where I wanted to be. I suspect this type of activity is more common than we might think. Some time ago, comScore did a study looking at search behavior as people progress through the buying funnel. Somewhat to their surprise, they found that often queries didn’t change, even as people moved down the buying funnel closer and closer to a purchase. Common sense seems to indicate that as we move through the consideration set and start focusing on a few key brands, our searches should become longer and more specific. In fact, comScore found that quite often exactly the same query is used through the buying cycle, right from early awareness through to purchase. Perhaps one of the reasons for this is this retracing of steps based on the route knowledge we acquire as we do our online research.
The n dimensions of online
So, if we use landmarks and memorize routes to navigate online, how come we never create a cognitive map? How come we never move to the highest level of spatial cognition, survey knowledge? As I said before, it’s hard to create a cognitive map when there’s no physical framework used from which to develop that map. One of the ongoing challenges of trying to graphically represent the Internet is how to map it. The most realistic representation of the Internet probably sits within the algorithms of the search engines. Whether you’re looking at Google’s concept of PageRank and the linking structure of the web, or one of the other engines variations, the structure of the web hangs on the concept of relevance. And this brings us to the key difference between our online and off-line worlds. The Internet is n-dimensional. There is no width or breadth or depth. Even if you consider time the fourth dimension, time has little meaning online. The only dimension that matters online is how relevant content is to our need at that time. This is why search engines have emerged as such a powerful force online. Search engines act as the connector between our need, our intent, and all the relevant information that may be out there. The more successful the search engine is in fulfilling this role, the more successful the user experience is. While physical proximity is an important concept in real-world navigation, online it’s all about informational proximity. We give the search engine a clue about what we’re looking for and the search engine goes out and pulls dozens or hundreds or thousands of threads together for us, hoping that at the end of each of those threads is just the content we’re looking for. We can navigate down any of these threads with a simple click of our mouse.
But it’s not just on search engines that informational proximity is important. This fundamental online navigational behavior provides us a clue about what many sites should strive to achieve online. If a site can become a recognizable landmark in one particular online environment, it can virtually guarantee success in many different ways. It will become a navigational hub and a trusted source of informational proximity. The challenge for these landmarks is to be able to identify user intent and respond to it by pulling together threads, much as a search engine does, giving users the opportunity to navigate quickly to just the information they’re looking for. These threads may come from other pages within the site or from external resources. The key is to provide informational proximity and successfully interpret the need or desire of the user, bringing valuable navigation options together in an easily identifiable way, rich with information scent, and establishing the site’s place as a prominent landmark within its online environment.
This brings into sharper focus the need for user research and usability testing. First, we need to have a deep understanding of what it is our target user wants to do. The better the landmark site fills this need, the more it will be used as a hub. Google excelled as a quick, stripped down, task-focused search tool, with “get in and get out” functionality. It didn’t need to be sticky, because that didn’t match the intent of the user. But that leaves large areas of search undeserved; areas like discovery search, where we’re looking for a stickier landmark, or vertical search, where we need more functionality to interact and compare the results. Right now, I think users are looking for better alternatives in both these areas. Success will come from understanding how to be a better navigational landmark by giving users what they’re looking for.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.