Category Archives: Cognition

Satisficing, Bounded Rationality and Search

150px-HerbertSimonHerbert Simon came up with some pretty interesting concepts, among them satisficing, bounded rationality and chunking.

Before Simon, we commonly believed that humans came to optimal decisions in a rational manner, based on the information provided. We took all the data that was accessible, weighed pros and cons and used our cortexes to come to the best possible outcome.

Simon, in effect, said that this placed to high a load on us cognitively. In many cases, there was simply too much information available, so we had to make choices based more on heuristics, cutting the available information down to a more manageable level. He called this “satisficing”, a blend of satisfy and suffice. And Simon started saying this a half century ago. Imagine how this translates to the present time.

We have never had more information available. At the click of a mouse, we can access huge amounts of information. There’s simply no way we can process it all and come to rational decisions. And this brings us to another concept, that of bounded rationality. We’re more rational about some decisions than others. It depends on a number of factors, including risk, emotional enjoyment and brand self identification. Think of it as a chart with three axes. One axis is risk. We put more rational thought into decisions that expose us to greater risk. In consumer decisions, risk usually equates with cost, but in B to B decisions, it could also include professional reputation (related to but not always directly tied to cost). We’re going to put a lot more thought into the purchase of a car or house than that of a candy bar. Another axis is emotional enjoyment. This is a risk/reward mechanism to most decisions, and if the reward is one that is particularly appealing to us, we tend to be swayed more by emotion than rational decision. If we’re planning a holiday, we may make some irrational decisions (or at least, they might appear that way to an outsider) based on a sense of rewarding ourselves. We’ll treat ourselves to a few nights in a 5 star resort, when the 3 star resort would offer greater overall value. The final factor, and one that is usually buried somewhere in our subconscious, is how we use brands or products to define who we are. Now, no one usually admits to being defined by a brand, but we all are, to some extent. This touches on the cult-like devotees that some brands develop. Harley Davidson, Rolex, BMW, Apple and Nike all come to mind. Is a Rolex a rational choice? No. But a Rolex defines, to some extent, the person wearing it. It says something about the person.

Bounded rationality says that there are boundaries to the amount of rational thought that we can and we want to put into decisions. The amount we decide is sufficient depends on the three facts discussed.

Now, the use of Search tends to plot somewhere along this 3 dimensional chart. If risk is high and brand identification is low (buying software for the company), there is a high likelihood that search will be used extensively. If risk is low and brand identification is high (i.e. buying a soft drink or a beer) there is almost no likelihood that search will be used. In this case, the two factors usually work inversely to each other. Emotional enjoyment isn’t as directly tied to search activity. We will do as much (or as little) searching for a purchase that will give us great enjoyment as for those that won’t.

It’s interesting to watch how these factors impact search intent and behavior. Satisficing leads to a classic sort of search behavior, what I call I category search, where we use fairly generic, non branded queries that broadly define the category we’re looking at. Let me give you an example. Tomorrow my wife and I are headed to Europe for a week. We’re going to spend a few days in Portugal, then fly up to London for SMX (where I’ll be talking more about these ideas in some of my sessions). We’re flying into Lisbon, then renting a car and driving down to the Algarve region. I have GPS navigation software for my PDA, but only for North America. I wanted to get European software, but because of the limited use of it, I didn’t want to spend too much. The developer of my North American software didn’t make a EU version, so I turned to search to find a suitable candidate. Here there was no brand identification, some degree of risk (if it didn’t work in Europe, I’d be lost, literally) and no emotional enjoyment factor. My first search was what I call a “landmark” search. I wanted to find some sites to plot the landscape. Sites that listed and compared my alternatives would be ideal matches to my intent.

I searched for “pocket pc gps software”, knowing that “gps software” would be too broad. I soon found the sites were pretty much all about North American versions. Few of them offered or reviewed European versions. I spent several minutes on the TomTom site trying to order a European version from Canada but to no avail. Apparently TomTom doesn’t believe people in North America would ever choose to drive in Europe.

In classic “satisficing” behavior, I wanted to cut my research workload by setting some basic eligibility criteria: it had to work on a Pocket PC, it had to be reasonably priced (under $100 preferably) and it had to offer coverage for all of Europe (we’re going back to France and Italy next year and I’d like to use it then as well). My next search was for “pocket pc gps software europe”. This gave me what I needed to begin to create my satisficed list. Ideally, we want 3 or 4 alternatives to compare. I did find the TomTom choice, but I was already frustrated with this, and the price was over my threshold. Destinator also offered an alternative that seemed to be a little better match. It matched all the criteria, appeared to have some decent reviews and was available on eBay for about $75, including shipping. Sold! Was it the optimal choice? Maybe not. If I had spent hours more doing research, I could have probably found a better package or a better value. But it was good enough.

Chunking has to do with cognitive channel capacity, and the amount of information we can store in our heads, accessible for use. Again, we tend to maximize the available slots by creating chunks of information, grouping similar types of information together.

When you look at Simon’s work, even though the majority of it far preceded search engines, it sheds a lot of light on how we use search in a number of cases. If you want to tap into user intent, I would recommend finding out more about bounded rationality and satisficing. Chunking is probably worth a look as well.

Are Our Brains being Rewired?

I have to start out by thanking Nico Brooks and Jess Gao. Without intending to, they both provided me more than enough fodder for a rather lengthy column in Seach Engine Land on Friday.

Nico is the Chief Search Strategist at Atlas. Jess is our intern at Enquiro, who’s currently working towards her doctorate, specializing in cognitive psychology. Through different paths, they both gave me some major brain melting ideas to chew over. I’m still digesting, but you can catch the thought process in action on my column.

But consider this. What if our brains are being rewired by the internet? Some of our behaviors are innate. They’re our OEM operating software, put there by the manufacturer. Flight or fight. The need to procreate. The appreciation of beauty. This stuff is hardwired.

But some of our behaviors are learned. We’ve developed them as we go. The things sit in our temporary memory caches, and we can adjust them if they’re no longer working. The thing that started all this was how we learn to navigate a physical environment. First we look for landmarks, then we memorize routes, then we put the two together to create a cognitive map. Nico’s suspicion (and Nico, I hope I’m capturing the essence of the idea accurately) is that our need to identify landmarks and even our ability to memorize routes is probably innate. It’s just how we are programmed to get around. But cognitive mapping, at least in the essentially rectangular grid pattern that is common in the Cartesian coordinate model, is a learned behavior. Rectangles have no place in the n dimensional space of online, so as we spend more time navigating online, will we change our mapping process?

Then, with Jess, we had a great chat about how we perceive things, especially ads. There’s a great introduction to selective perception that I would urge you to check out. In recent studies we’ve done at Enquiro, one of the interesting findings has been that the more intrusive the ad, the less it seems to work. It registers high in the first stage of perception, stimulation, and manages to succeed in the second, registration, but fails in the last two stages, organization and interpretation.

Other conversations I had this week, that didn’t make it into either of the columns. On Thursday I was in New York for Google’s B to B Summit and had a chance to chat with Mark Martel, who supports the B to B Tech Sales Vertical at Google. Mark has a healthy intellectual curiosity and I always enjoy chatting with him. We discussed schemas and how important they are in the process of perception. Then, on Friday, I was in Toronto chatting with the Yahoo Canada gang, including Maor Daniels and Adina Zointz (what a great name, literally covering everything from A to Z!) and we talked about how quickly we’re learning to judge the authenticity of content online. It’s as if our bullshit filters are more finely tuned than ever.

I’m definitely on a riff here, but there’s a lot of threads coming together. Even in someone of my ever upwards creeping years (I’m 46) I suspect my synapses are under construction. Old routes are being torn out and new ones are being built. And with my daughters, many of the paths are being built differently right from the start. The routes that were so important to me in grade school, times tables, rote memorization, etc, are becoming overgrown with weeds through lack of use. But new routes I never even thought of, like how to do homework, carry on an online chat and watch the TV with one eye, are being upgraded into major turnpikes. Multitasking is a major operational imperative now, and selective perception is kicking into overdrive.

Anyway, to further dive into some of the things on my mind, here’s some of the columns where I’m beginning to open up some of these ideas to the fresh, online air:

Infomediating a Broken Marketplace – a look at Hagel and Singers Infomediary model from their book Net Worth. Is Google aiming to be the ultimate match maker in the marketplace?

4000 Ads a Day, and Counting – Part One of the Infomediary Doubleheader, looking at the disconnect between customers who just want the facts, and advertisers that just want to control our buying habits

Some Big Ideas for a Friday – Some musings about how we perceive advertising, based on recent studies we conducted, and how we might be remapping the perception process

How We Navigate Our Online Landscape – The original exploration of landmark, route and survey knowledge and how it may map (or not) to how we navigate our online space

And please, do me a favor. This is all stuff I want to explore further in the book. If you think I’m full of bullshit, call me on it. Share your thoughts. Post a comment. Start a dialogue. I know it’s a pain in the ass posting comments on blogs because of spam, but PLEEASSSE take a few moments to do so. Or drop me an email.

The Ultimate Market Research Technique?

sharingbrainThis is kind of cool, in a really creepy way. According to a recent study, Scientists can now tap into the brain and predict whether you’re going to buy something or not. Not to get all scientific on you, but apparently a portion of the brain called the nucleus accumbens “lights up” on a brain scan if you’re ready to whip out the plastic. But, if the price tag is out of your budget range, another region of the brain called the insula is activated and the mesial prefrontal cortex is deactivated. Dr. Brian Knutson of Stanford and his team are doing the research.

So, think of this future scenario:

Google gets wind of this and brings this into the Google Labs. They work with Intel to develop a small implantable chip that constantly monitors this part of the brain. Through a secret agreement with the U.S. Government, giving the Homeland Security teamaccess to everyone’s online history, Google gets the right to implant the chip in every new child born in the U.S. The chips are connected through wi-fi, so that Google can monitor everyone’s inclination to make a purchase. You can now test your Google campaigns right down to the purchase, setting up A/B tests with the ultimate feedback loop.

Mmm..the mind boggles with the possibilities here….

The 50 Millisecond Judgement

Originally published February 2, 2006 in Mediapost’s Search Insider

Fifty milliseconds is not a long time. It’s about one frame of video, or half as long as the blink of a human eye. And that tiny little slice of time is all it takes for a visitor to a Web site to decide how appealing that site is.

Dr. Gitte Lindgaard and her team undertook a fascinating study at Carleton University in Ottawa, Canada. Their goal: to determine just how long it takes to make a reliable judgment of the visual appeal of a Web site. They found that we can accurately judge visual appeal in just 50 milliseconds, or one twentieth of a second! The study was published in Behaviour and Information Technology, March – April 2006.

In three different studies, the Carleton team flashed home pages of Web sites, specifically chosen to provide a spectrum of visual appeal, at participants for varying lengths of time and then asked the participants to rate the pages from 1 (very unappealing) to 100 (very appealing). In the first two studies, the duration of exposure was 500 milliseconds, or half a second. In the third study, participants were randomly shown the pages for either 500 milliseconds, or 50 milliseconds. The ratings were then correlated and analyzed to determine the reliability of the rankings. Dr. Lindgaard’s team found that reliable assessments of visual appeal can be made even with a 50 millisecond exposure.

Beyond this finding, however, there were a number of topics touched on in the paper that site designers should take to heart. While these topics weren’t included in the scope of the study, the paper cites numerous studies that have tried to explore the nebulous area of visual attraction and how we determine it.

Blink Revisited

For anyone who’s read Malcolm Gladwell’s book Blink, you know that researchers are discovering that humans make decisions in two very distinct ways. On a visceral level, it seems that a decision-making mechanism is hardwired right into our physiology. Our bodies seem to reach conclusions long before our brain catches up. To quote Dr. Lindgaard’s paper, “More recent neurophysiological evidence supports the contention that emotional responses can indeed occur pre-attentively, before the organism has had a chance cognitively to analyze or evaluate the incoming stimulus or stimuli. A small bundle of neurons has been identified that lead directly from the thalamus to the amygdala across a single synapse, allowing the amygdala to receive direct inputs from the sensory organs and initiate a response before the stimuli have been interpreted by the neocortex.”

After this very brief response, we begin to rationalize our response by logically evaluating the stimuli. The two-phase decision-making mechanism typically works together to help us reach our conclusions. Gladwell’s contention is that the first response, the “blink” response, is often the right one.

First Impressions Do Count

So, how important is that first, split-second decision in online interactions? Because of something called a “halo effect,” it can be vital. If we have a positive emotional response in those first few milliseconds, our logical mind will kick in and try to rationalize that response. We will look for positive reasons why it was the right decision, and we will tend to ignore negative factors. If the first impression is not good, the opposite occurs. We look for reasons not to like something, and tend to discount any positives we might find. We want to prove our first impression right.

Translated to an online experience, we make an immediate, intuitive decision whether we like a site or not, without reading one word of content. From that moment on, our entire interaction with that site is colored by that first impression.

Is Beauty in the Eye of the Beholder?

How do you judge what is appealing or what is not? It’s a thorny issue, as Dr. Lindgaard acknowledges in the paper. It’s been said that we all have different concepts of what’s beautiful or appealing. However, there was remarkable consistency across all three studies with the sites that were found appealing, and the ones that weren’t. In fact, it was found that in groups of as small as 5 people drawn randomly from the larger group, consensus emerged on the winners and the losers. So while beauty may be in the eye of the beholder, it appears that we pretty much see eye to eye when it comes to Web sites.

One reason might be in the factors we use to judge appeal on a Web site. We are not looking at it as a pure object of aesthetic beauty. A Web site should be usable, so we are also making an assessment of how appealing a site would be to use. We’re looking for a site to be clean, pleasant and symmetrical. We’re looking for proper use of screen real estate and balance. Previous research by Dr. Marc Hassenzahl suggests that we may use two methods of evaluation, which he refers to as beauty (the pure aesthetic appeal) and goodness (the more practical factors, including usability).

One thing that wasn’t covered in the study was seeing how sites rank when user intent is added to the mix. The participants in the study had no particular goal in mind. They weren’t looking for anything. I would love to see what happens when we introduce intent and participants are judging sites not just on appeal, but on the promise of delivering on their intent.

A Qualitative Research Primer

For anyone who’s interested in qualitative research, this study offers some valuable tidbits on testing methodologies. It’s an interesting challenge to gather results on something as raw and intuitive as a first impression. The minute you start to analyze the response, you distance yourself from that first visceral reaction. Does the very act of rating a site kick in the rational mechanism and bias the original response? As in most studies, Dr. Lindgaard acknowledges that there are many more questions to be answered here. In a brief chat I had with her, she expressed her eagerness to continue down this path, “This is just the tip of the iceberg,” she said. “There’s so much more here!”

Back To Those 50 Milliseconds

What does this have to do with search marketing? Well, everything. Through the utilization of search engines, you will hopefully be driving thousands of new visitors to your site. That’s thousands of first impressions, formed in less than the blink of an eye. Search marketing is useless if it doesn’t deliver a positive onsite experience. Our obsession with position and click-through is meaningless if all our efforts (and all budgets) are blown apart by those first 50 milliseconds.

It is my intense belief that the key to success lies in better understanding what happens when those synapses fire and those first impressions are formed. It’s not just an understanding of the mechanics of the Web that will create a successful search marketer. It also helps to peer into the awesome machinery of the human mind.