Human Hardware: The Subconscious Side of Search

In the last column, we explored some studies that indicate that conscious will might not be the cause of our actions, but instead might be just one of the effects of a motivator still undefined. Conscious will could simply be a feedback mechanism that helps us keep track of our actions and hides the degree of unconscious activity involved in our day to day activities. Conscious will allows us to keep thinking we’re rational, logical beings.

Energy conservation of the brain

Today, let’s look at how that might play out in a search interaction and other online behavior. As I said last week, habits and conditioned behaviors are the brain’s way of conserving energy. If we continue to be faced with the same circumstances and our actions provide the same response, we stop thinking about doing those things consciously and relegate them to unconscious preprogrammed scripts. These are what we come to know as habits.

Many of our activities online have become habits. We have our favorite starting places, resources and routes through the web. We don’t have to think about them. We’ve been here before, we know what to expect and we know how to navigate our way through these resources. We can cruise through on autopilot, and that sense of familiarity is a powerful draw for us. Researchers (Murray, Hauble, amongst others) have coined this “cognitive lock in”. Steve Krug called his book on website usability “Don’t Make Me Think” and he gets it absolutely right. Our first choice is to navigate by habit. Our second choice is to navigate by instinct. Either of these can be done with a relatively small degree of conscious interaction, and we naturally prefer this. If we have to think, we have to use brain power, and we’d rather not do that if we don’t have to.

The forming of habits

Habits form through repetition. The more times you do things, with reasonably consistent results, the more likely it is to become a habit. And once something becomes a habit, it’s very difficult to change behaviors. The more something is repeated, the deeper the habitual rut we carve. There is no online activity (with the possible exception of use of email) that is done more often than searching. And there we have the crux of the market share dilemma for every engine that isn’t Google.

One of the biggest “why” questions in search is why Google has been able to build huge market share without advertising, while Microsoft, Ask and previously, NBC and AltaVista, threw huge amounts at advertising without moving the market share needle one iota. In fact, in most cases, share eroded after the advertising. The answer? Google has become a habit. We don’t think about using Google. We just do it. And for advertising to be effective, you have to make a conscious decision between alternatives. For most of us, this decision never takes place. Or, as per the last column, we have the illusion that we consciously choose Google when in reality, the action was initiated unconsciously and it’s only that deceitful little “conscious will” feedback loop that leads us to think that any rational decision was involved.

How Google became a habit

Let’s look at how I believe Google became such an ingrained habit, and it will help to explain how these programmed scripts get written in the first place. Pre-Google, our searching was not entrenched with any one property. Personally, I favored AltaVista, or, before that, Infoseek, but the success rate was so low that I often switched my searches to Lycos, Excite or Yahoo! I found that for certain searches, some engines offered better odds than others. So my search activity was spread over a number of properties, depending on my intent, the nature of my search and sometimes, where my online starting point was (i.e. my home page). None of these properties ever became a habit, because none of them delivered consistent enough quality to keep me from going elsewhere.

Then, along comes Google. While Google undeniably had better relevancy technology, there was also a large degree of luck in timing as well. First, when the PageRank concept was applied to the existing corpus of web information, it did yield better results for more searches than any other engine. Out of the gate, Google was a superior experience for more searches, curbing our need to go elsewhere.

But luck also played an important but often uncredited role. It’s not the actual relevancy that’s important to the user, it’s the perceived relevancy, and it’s here that Google had the advantage of two huge helping factors.

First, the early Google is a text book example of viral growth online. Use of Google, even in beta form, started in the academic and journalistic communities. In the spread of influence through word of mouth, we defer to perceived authorities, and who could be more authoritative in the seeking out of information that academics and journalists? Google came with the information geek’s stamp of approval firmly affixed. True, Google offered a better search experience, but the actual value of this was compounded by the nature of the evangelists who took up the Google banner.

Timing is everything

Viral growth wasn’t the only factor. Timing also played a huge role. At the same time as Google was growing virally through word of mouth, there was also an explosion of web content, as the Internet hit critical mass. According to Netcraft, in April 1997, there were 1 million websites. This was just as Google was evolving from Sergey and Larry’s dorm room project to a commercial entity. Over the next 4 years, Google would move from the beta, word of mouth darling to search megalith. In that same period, web content exploded by a factor of 10, growing to almost 10 million sites by 2001 (by the way, we just passed the 100 million site mark, so another 10X milestone has been reached).

Let’s consider the interplay between these two factors. Just as Google was quietly winning new users because of the promise of more relevant results, the amount of content they could draw from was also exploding. High relevancy was a much surer bet, because there was a bigger well to draw from.

All these factors combined to establish Google as the search engine of choice. And the more time we spent on the Google interface the more we didn’t have to think about it. Google became a habit.

Breaking habits

This presents the dilemma facing Google’s competition: how do you break a habit? Reams of academic research (Verplanken and Wood, 2006, for one), especially in the healthcare sector, shows that it’s extremely difficult to break a habit by rational appeals after the fact. You can implore people not to smoke, drink to excess, or eat unhealthy foods, but to little avail. In fact, many of these campaigns have actually been proven to stimulate the unwanted behavior. The visual cue kicks an unconscious craving into action and, literally before people know what’s going on, they find themselves lighting up or scarfing down a Butterfinger bar. It’s not a question of will power, it’s the simple fact that the action has already been set in motion before you ever have a chance to rationally consider it. All the will power in the world will simply close the barn door after the horse has already bolted.

The challenge with habits is that we don’t think before we do them. The unconscious behavior is what defines a habit. They’re a cognitive short cut. It’s not like we spend minutes, or even seconds, deliberating on whether we’re going to use Google or Yahoo! for our next search. We’re merrily Googling away before our rational brain ever checks in. So all the advertising in the world, or, for that matter, Microsoft offering cash search kickbacks, will likely prove unsuccessful in encouraging breaks with habitual behavior. You’re walking into the fight armed with the wrong weapon. To break a habit, you have to move upstream from the habitual behavior and disrupt it. A dieter will have much more success if they clear their environment of any fattening foods, rather than relying on willpower alone to keep them from reaching for the Haagen Daz ice cream bar. Similarly, a search engine will have more success if it can disrupt the Google search habit before the search is launched. This, by the way, is the only reason Yahoo! has hung on to the search share they have. Right now, the Yahoo! search box is still present in enough crucial online search intersections to capture the query before it goes to Google. Microsoft used to have this advantage with MSN, but ironically, they broke their own habit by rebranding to Live Search, much to the detriment of their search share. They’re currently repeating this mistake with the drastic interface changes unveiled in Office 2007. The impact will be less here, because at this point, there’s not a strong viable alternative, but they’re still forcing people to think their way through a new interface, eroding the habitual advantage they had.

Habitual searching

So, the habitual choice of a search engine is one area where the illusion of conscious will is at play. The actual interaction is another area. The typical search interaction is a classic example of scanning presented information in the most efficient possible way to yield the optimum result. And, through conditioning over thousands of search experiences, we’ve refined our behavior into some repeatable, somewhat unconscious behaviors.

Take query structure, for example. Search providers have constantly bemoaned the lack of sophistication in query development. We almost never use advanced search capabilities, our queries are strung together quickly and we try to use the fewest words possible. It’s a quick mental burp, targeted at the query box. We don’t want to expend too much cognitive effort in something that should be fast and intuitive. Searching should be a natural extension of the rapid firing of our neuronal synapses, not an arduous and deliberate process.

Now let’s look at the search interaction itself. This is incredible rapid. A typical search interaction is about 10 seconds (before the first click) in which 4 to 5 results are scanned. That’s about 2 seconds per result, which is not enough time to deliberate about anything. In many people’s minds, they have the illusion that they’re carefully considering all the results, sifting through them to find the best match to intent. In reality, they start in the upper left, scan down quickly looking for a few relevancy cues (the major one being the query they just used being bolded in the title of the listing) and if there’s enough information scent, people click. This type of interaction, while conscious, is skating on the edge of our unconscious as well. As we scan information on the page, there’s a stream of words and cues going through our conscious executive minds, each of which may unlock a corresponding concept buried at an unconscious level that colors our interaction with the page. For example, a search for “SUV” might include brands that have a corresponding meaning for us, either good or bad. A promise of model comparisons might resonate with our intent to find and consider options. A snippet that talks about fuel efficiency might connect with a desired feature or even a more altruistic concern about the environment. In our previous research, I’ve referred to this as semantic mapping, and although we’re consciously engaged with the task of searching, there’s a tremendous amount of processing that’s happening at an unconscious level, helping us determine in the space of a few seconds which link we’ll click on.

Judging in the blink of an eye

Finally, when we do click through to a destination page, we make immediate emotional and aesthetic judgments about the quality of the site. It’s been shown that we can judge the visual appeal of a site in as little as 50 milliseconds (Lindgaard, Fernandes, Dudek and Brown, 2006). That, by the way, is about half as long as it typically takes you to blink your eye. Incredibly, these researchers found that we can accurately assess aesthetics in this tiny slice of time. So, before you ever have the chance to determine the value of content, you’ve already passed judgment on the appeal of the page. In psychological terms, this primes the rest of your interaction, either negatively or positively. And, again, this happens at a unconscious level.

So, from the second you start to search to the time you find the site, there’s a corresponding stream of unconscious behaviors that color and influence your conscious decisions. The two streams are mixed and masked in our minds, leading us to believe we are conducting ourselves in a purely rational and conscious manner. And, if I stopped you afterwards and asked you about your search experience, you’d automatically give me perfectly rational reasons why you did what you did.

One of the misleading things about search is that it looks deceptively simple on the surface. But like most human activities, there’s a lot happening just below that surface.