This is part two of my series on expectations and usability. I have to apologize for originally saying this entry would be about standards. By "standards", I meant standard practices, not web standards (W3C, etc.), so, to avoid confusion, I've decided to use a broader word: habits. When we browse the web, many habits come into play, and just about all of them influence our expectations.

What Are Habits?

As a cognitive psychologist, I use the word "habit" pretty loosely, to refer to just about any learned pattern of behavior. A habit is generally learned through repetition and may be influenced by industry standards, cultural standards, technological and environmental restrictions, and even personal preferences.

Biology vs. Habits

Some of our expectations are essentially built into our biology. We tend to perceive the color red or large text as being more important, for example. On the other hand, our reaction to something like italicized text represents a habit. We've learned to accept certain publishing standards in written text, including the idea that italics indicate emphasis. Of course, the line between biology and habits isn't always black and white, but to understand expectations, it's important to recognize the difference between how we're hard-wired and what we learn along the way.

Two Types of Habits

For web users, we can roughly split habits into two types: (1) Cognitive habits, and (2) Industry standards. Italicized text indicating emphasis is a cognitive habit: we generalize it to reading text in just about any modality, including books, newspapers, the internet, and our mobile phones. Cognitive habits often have a cultural component, such as English being read from left-to-right. Industry standards drive habits specific to web browsing, and are a consequence of standard practices over the past 10-20 years. Blue, underlined links are an industry standard. Interestingly, the worldwide nature of the web means that industry standards are often cross-cultural, a fact which is slowly changing expectations for global users.

Example: Banner Blindness

To understand how habits affect expectations (and with them, usability), a couple of examples would probably be more useful than more explanations. One phenomenon that has become well documented in the usability world is that of banner blindness. Put simply, people ignore web-page elements that look like banner ads. This is entirely a learned habit, and one based on industry standards. A 468x60 pixel rectangle has no intrinsic or cultural meaning for us, but we've learned to expect it to be an ad, and so we ignore it. This not only drives expectations, but demonstrates just how quickly a habit can be learned by a large population of users. In less than a decade, banner blindness is a virtually automatic response for most of us.

Example: F-shaped Pattern

Another well-researched habit is the so-called F-shaped browsing pattern (see image above). Although this habit has some cultural components (such as the Western tendency to read left-to-right and top-to-bottom), it has also been driven by web design standards, which have tended to place critical content near the top and left edges of a website. Over a relatively short-time, this has driven our expectations, causing us to automatically move our eyes in an F-shaped pattern over a new site.

Habits & Expectations

Whether our habits are general or specific to browsing the web, these habits drive our expectations. If, as developers, we ignore these habits, we risk violating people's expectations and creating mental roadblocks that our visitors may find difficult to cross. Understanding habits is key to understanding and meeting user expectations.

I have to apologize: I completely forgot to mention on the blog that I'd be gone for 10 days in Taiwan. Had a great trip and just got back last night. I'll be continuing my series on expectations and usability next Tuesday. In the meantime, here's one of my 250 pictures from the trip: sunrise over the Alishan mountains.

In a surprising move today, Google announced that Vice President Dick Cheney will be joining the company in January of 2009 as Vice President of Global Affairs. Cheney announced his new role at a press conference this afternoon:

I'm pleased to announce that, when my term ends as Vice President, I will be joining the Google team. Google is uniquely poised to tackle a problem that has been important to Lynne and I for years, that of global domination.

Effective immediately, Google will change it's long-standing "Do no evil" motto to "Do some evil". As Cheney explains, "For too long, Google has had its hands tied by an irresponsible anti-evil philosophy. In this post-9/11 world, we can no longer afford to take any form of evil or mad science off the table."

After the press conference, Google co-founder Sergey Brin expressed his optimism about the newly forged alliance:

Larry and I are very happy to have Dick on board. At first, some of his ideas seemed unorthodox, but we've finally begun to realize that many of our initiatives, such as plug-in hybrid cars, simply aren't practical. Thanks to Dick, we'll be exploring projects with much greater real-word potential, such as moon lasers.

Brin admitted that some of Cheney's ideas were still being met with resistance: "Dick's suggestion that I change my name to Sam because Sergey is 'kind of Commie-sounding' is still under review, but I think I'll come around."

To Sergey, Larry, my readers, and the search industry, my sincere apologies. To Mr. Cheney, my sincere hopes for your rapid transition to the private sector.

I've been reviewing a lot of small business websites lately, and as I see the same usability issues popping up over and over, it occurs to me just how many of those problems are a result of violating visitors' expectations. So, I've decided to write a short series (3-4 installments) on the role of expectations in usability, and how we can better anticipate those expectations to help website visitors accomplish their goals.

What Are Expectations?

It may seem like a silly question, but what exactly are expectations? Essentially, expectations are the unconscious predictions we humans constantly make based on our models of the world around us. In other words, we use our past knowledge and experiences to predict the near future. Expectations are, in many ways, evolution's ultimate survival mechanism.

Cognitivism & Mental Maps

The man in the picture is one of my professional heroes, Edward C. Tolman, early 20th-century psychologist and one of the fathers of the cognitive revolution. It may surprise you to know that the idea that people can make plans and build internal models of the world is a relatively new one, and met with intense skepticism throughout much of the 20th century.

The cognitive revolution ushered in the idea that animals and people can create mental models or cognitive "maps". Prior to this, it was assumed that all behavior was a function of complex stimulus/response mechanisms. In other words, our behavior was determined directly as a result of outside events triggering learned patterns. In the behaviorist model, we were victims of our past and environment. The cognitive model started to see us as proactive, creating models and expectations and anticipating instead of merely reacting.

Violating Expectations

In the internet world, we expect, from our experiences, that websites will follow certain standards and be self-consistent. So, what happens when those expectations are violated? In essence, the brain does a double-take, forcing people to stop (even if for a split second) and re-evaluate. With every violation, the time needed to process information grows and frustration increases.

In daily life, we encounter these double-takes without even realizing it, and probably make thousands of adjustments to our mental maps every day. Eventually, though, these expectation violations add up, and we look for alternatives that match our mental map. On the web, where alternatives are a click and mere seconds away, frustrations build very quickly and violating expectations can have immediate consequences. The lack of patience of online visitors makes meeting expectations especially important for commercial websites.

Types of Expectations

Over the next three installments, I'm going to talk about how website visitors expectations are determined by different factors: (1) standards, (2) consistency, and (3) search behavior. With each, I'll discuss how we can better meet these expectations.

If you've ever seen a kung-fu movie, you're probably familiar with this story: kung-fu master has a secret technique, master trains worthy student, student uses technique to gratuitously beat everyone to a pulp. The idea of a secret technique known only to a selected few has always been a popular one, promising fame and glory to whoever can unlock that secret.

Unfortunately, I see this same mentality too often in the way people run their websites and e-businesses. Otherwise rational people desperately want to believe that there are usability, SEO, and marketing secrets, and if they can just get the so-called experts to reveal those secrets, the money will flow like water.

Secrets Are Good Marketing

Secrets sell. Even the kung-fu cliché is based on real history: martial arts schools were often businesses competing for wealth and power, and protecting their "secrets" from the masses increased their allure and demand for new students. In much the same way, web professionals have learned that dangling a secret recipe in front of a prospective client can be much more lucrative than trying to explain difficult concepts and long-term strategies.

The Bullet Point Mentality

Unfortunately, we're all a bit lazy, and we tend to adopt a bullet-point mentality about things we don't understand. In other words: don't bother me with all of the gory details, just give me the bullet points. We flock to "Top 10" blog posts and "101 Tips" articles, hoping to find the secret sauce poured over a few bite-sized pieces. There's nothing wrong with trying to simplify difficult concepts, but when we reduce everything to a bulleted list, the end result can become so over-generalized and ambiguous that it's impossible to effectively apply to any given situation.

It Only Looks Like Magic

Another problem is that experience has a way of making things look easy. This is one of the hardest parts of being a consultant, in my experience. A skilled usability or SEO specialist can look at a site and spot many major problems in less than 15 minutes, leaving clients thinking that what we do is easy or just the result of some secret we aren't sharing. The reality, though, is that our gut reactions and ability to spot problems is often a result of seeing similar problems hundreds of times.

Solutions Require Effort

Perhaps the biggest problem with a secrets mentality, though, is that it really only focuses on the problem. I could probably teach someone how to spot a handful of the most common website usability problems, even with limited experience. Solving those problems, though, often requires research, testing, and a lot of hard work. Kung-fu movies conveniently skip over years of training, much like Rocky learns to box in the space of a video montage. Real life, on the other hand, can't be put on fast-forward.

Don't worry; this isn't a post about how I found religion. I've just been thinking a lot lately about conversion (the website performance metric, that is) and how we've come to regard it as the Holy Grail of metrics. Now, don't get me wrong: measuring conversion is incredibly useful, and has been a huge boon to website testing and optimization. The problem is, we're starting to take it a bit for granted, in exactly the same way we have every metric before it.

A Brief History of Metrics

A very, very brief history. If you look at website metrics in general, you'll see an evolution in the past decade or so that goes something like this:

Hits > Page Views > Visitors > Unique Visitors > Conversions

Interestingly, the online advertising industry has seen a similar evolution:

Impressions > Click-throughs (CTR) > Conversions

In each case, we realized that something about the previous metric was misleading or simply lacking. Ironically, though, in each case we started to view the new favored metric as being the perfect measurement (until someone showed us a better one).

Limitations of Conversion

So, how is conversion lacking? First of all, like any metric, conversion is relative and depends on how you measure it. Is a 10% conversion rate good? Not if your conversion was 15% last year. More importantly, it doesn't reveal the full picture. Like all of the metrics before it, improving conversion is generally a good thing, but not necessarily.

Take this example: let's say you have an e-commerce site and you sell Product A and Product B. Currently, your site gets a 3% conversion rate. You decide to test a major change to the site and your results show a 6% conversion rate. Great news, right? Maybe not. What if Product A has a profit margin of $1, and Product B has a margin of $100. Prior to testing, 90% of the conversions were purchases of Product B. The new version of your website completely shifted the balance, and Product A accounts for 90% of the purchases. Your conversion doubled, but you just lost a lot of money.

What Comes Next?

So, the obvious question is: what comes after conversion? First, let me ask another question: are all of the old metrics useless? Of course not. Let's look at online advertising; having a high conversion rate is great (all else being equal), but you still need impressions to drive click-throughs, and click-throughs to drive conversion. It's important for metrics to evolve and improve, but we can't ignore everything that came before the latest and greatest.

In the end, no one metric is ever going to tell the whole story, and we have to let go of conversion's Holy Grail status long enough to see the big picture. Start by considering these questions:

• How does your conversion rate compare to the past?
• How does your conversion rate compare to your industry?
• Are you measuring the right goal?
• Should you be measuring multiple goals?
• As one goal improves, does it harm another goal?

Note to self: Apparently, 250 subscribers isn't quite enough to start running contests :) Either that, or I scared you all off with my stipulation of posting the winners' reviews online.

At any rate, thanks to my 4 entrants, and since you all seem to be legitimate, you're all winners! I'll be contacting you personally over the next couple of days. I've decided that, since the point of this was mostly to test-drive some new service offerings and some entrants submitted on behalf of clients, I'll be keeping the reviews private.

Thanks to those that entered, and I'll be back to my normal blog posts next week.