Read the full article at The Results People blog.

I sure hope that you included a URL in your ad.

Whether you like it or not, potential customers who watch/read your ad will go looking for you online. Are you making it easy for them to do that?

In a research study that we recently did for one of our clients, we showed participants a TV commercial. In an overlay at the bottom of the frame, each participant saw one of three things:

• The company name (no call to action)
• A phone number
• A vanity URL

After watching the commercial, we asked each participant what they’d do to follow up (they were allowed to pick more than one activity).

Visiting the website and searching for the company were the top two activities, regardless of the call to action.

Hang on a second, you say. According to that chart, participants were more likely to visit the website when there was no call to action. Good point.

Except that when we asked them what website they’d visit, they told us the wrong one. The client was using a vanity URL that the participants couldn’t remember or guess – instead, they guessed at a URL that would have ended up taking them to the site’s homepage, where there was no information relevant to the commercial. If they’re looking for you online, wouldn’t you like them to find the information they’re looking for?

This led us to recommend that the client:

• Use URLs and a phone number in their commercials
• Use folders in their URLs (www.company.com/promo) rather than vanity URLs (www.buyfromcompany.com)
• Reinforce the call to action by including “call 1-800-555-1234 or visit company.com/promo” in the commercial voice over

So now that we’ve established that your potential customers are going to try and find you online, let’s take a look at how many of you are actually making that easy for them to do.

A recent study from Nominet looked at the use of URLs in print and TV advertising in the UK. The UK is pretty sophisticated when it comes to online stuff.

Only two thirds of print and TV advertising had a URL. When you break that apart, 83% of newspaper advertising had a URL, while only 61% of TV advertising did. In an analysis that looked at £40 million worth of advertising, that means that tens of millions of British pounds of advertising maybe aren’t being used as effectively as they could be.

Are you throwing money away by not including a URL in your advertising?

Do you use URLs in your offline advertising? Why or why not? Let us know what you think by commenting below.

We recently ran an analysis for one of our clients of the length of keyword query (whether it is one-word queries such as “dog” or two-word queries such as “dog catcher”, etc.) for the two weeks leading up to Google Instant compared to the two weeks prior. For this particular client, three-word search queries are the main driver of traffic to their site followed by two-word and four-word queries.

Here are the results:

We conducted a statistical analysis (thanks, Ian) to determine if the changes we saw pre and post Google Instant launch were actually statistically significant. Using a 95% confidence interval, only the change in one-word search queries could be considered statistically different (a decrease from before Google Instant to after). Interestingly, the vast majority of those one-word search queries were for branded terms. Given our knowledge of our client, we know that they released a new product in August which generated additional traffic to their site. The decrease in search volume could be attributed to falling off interest as people looking for their new product have already had the time to search it out and learn about it.

Our next areas for further analysis are:

- looking at clients of ours who are more dependent on one and two-word queries as a driver of traffic to their sites to see if people are increasing using Google auto-suggested search queries

- looking at the change in the number of impressions for client sites, which we’ll be doing for our clients once that data becomes available in Google Webmaster Tools.

If you haven’t read the book yet, The BuyerSphere Project investigates how business buys from business in a digital marketplace. Maybe that doesn’t sound like a big deal, and you’re immediately thinking of buzzwords like B2B, funnel, long cycle, enterprise procurement and so on. Well, to make a long story short, after all the research we put into the BuyerSphere Project we found a few things about B2B buying that, frankly, went against the common wisdom.

If there’s only one thing you get from the book it’s summed up in one word: risk. B2B buying behavior, from paper clips to heavy machinery, is influenced by the risk factor. And not just the institutional risk, but also the personal risk factors of all the individuals involved in the purchase process. As we dive deeper into this territory, things get complicated fast.

Chapter One introduces the important concept of mapping the BuyerSphere for your company. The reason you need to create a map is so that you can get a clear external picture of how the buyer views risk as it relates to your business. Once you have a clear map of the terrain between you and the buyer, you’re in a stronger position to plan your sales and marketing approach effectively.

The BuyerSphere has three dimensions: the product, the market, and the buyer. There are questions you need to explore in each of these dimensions that will help your map take shape. In the book, the three dimensions are shown as overlapping circles:

The intersection of the three circles (or spheres) defines your marketplace. But how do you map your company’s BuyerSphere? I have some ideas about that I’ll share with you next time. Until then, you might want to catch up on some of our BuyerSphere webinars and more.

• Recognizing a need for information
• Accepting the challenge to take action to fulfill the need
• Formulating the problem
• Expressing the information need in a search system
• Examination of the results
• Reformulation of the problem and its expression, and
• Use of the results

(Marchionini and White in Hearst, 2009)

An interesting criticism of this model came out last year in a paper by Max Wilson called “The Tetris Model of the Information Seeking Process”. The standard information seeking model sees users as formulating their queries, viewing their results, and reformulated their query ad infinitum as needed in a circular process until they reach their desired result. Wilson argues that the stages of information seeking don’t necessarily occur in a linear process but can be better visualized as a Tetris layout where information must be fitted together to reach a goal. In this model, progression, time, and movement between the different stages of information seeking are tracked independently. This model visualizes the quality and complexity of information through the depth of a Tetris block:

While Wilson relates the increasing speed of the Tetris game to the deadlines and time constraints that people face when searching, I see it as also representing the rate which users become familiar and proficient with the search interface. A game that doesn’t speed up could be seen as a metaphor for usability problems. While the model doesn’t assist with aspects of the information seeking process such as query formulation, it does provide new and novel way to visualize how we piece the information of multiple search results together to find an ultimate result.

In reading the study, my mind went back seven years to one of the first research studies Enquiro ever did (and still our most popular download) – Inside the Mind of the Searcher. In it, we observed the behaviors of 24 individuals as they used search engines to carry out tasks. It was the first qualitative study we did, before we used eye tracking. But the Wesleyan study brought back some interesting insights from that study.

As we looked at the group, we started seeing some different search strategies. In the paper, we divided them into four groups: The Scan and Clicker, The 2 Step Scanner, the Deliberate Research and the 1,2,3 Searcher.

Here is a brief description of each:

The Scan and Clicker (12.5% of our total group)

These people scanned the top 3 or 4 results and clicked right away if they found something of interest. They were less likely than the 2 Step Scanners to return to the results set.

2 Step Scanners (25% of our total group)

They’d scan the top results, same as the Scan and Clickers, and might click on a listing of interest, but would tend to “pogo stick” more, clicking through to a site, but then returning to the search engine and checking out at least one or two other sites before committing to one site.

Deliberate Researchers (41.6% of our total group)

This group felt they had to scan the entire results set before clicking on a result. This group spent the longest time on the page, almost 40 seconds, compared to 15 to 20 seconds average duration for the other searchers.

The 1,2,3 Searchers (20.8% of our total group)

This group worked down the results in order, seeming to consider each result individually. There didn’t appear to be as much back and forth consideration as we saw in other groups. Of course, we weren’t using eye tracking, so it was difficult to accurately track specific eye movements.

Now, these sessions were recorded 7 years ago now, so I suspect some of the behaviors we saw modified themselves as people became more familiar with search engines. I’ve talked before about how we develop conditioned strategies through repeated tasks. Search is a prime candidate for this.

Decisiveness and Search Patterns

But in looking at this, it does seem that the same decisiveness vs. indecisiveness behaviors identified in the Wesleyan study were also appearing in ours. One of the interesting things I’ve found in our own research, and something also alluded to in the Wesleyan study, is that you need to track behaviors in minute detail before you start to see the nuances that may indicate different underlying strategies. For example, we’ve seen aggregate heat maps that look almost identical between two groups, but it was only when we walked through the eye movements on a second by second (even a quarter second by quarter second) basis that we saw people taking significantly different paths to end up at the same destination. In the Wesleyan study, they found that under the pressure of time, indecisive people might abandon maximizing strategies to adopt satisficing behavior. This may yield similar results at the end, but can generate greater levels of stress and anxiety on the way to a decision.

I suspect decisiveness could be a critical factor in how we might navigate any web page, including a set of search results. For example, how would decisiveness impact our interaction with the sponsored ads at the top of the page, or visually richer results? Great questions, currently with few answers.

I’ll see what I can do about that.

Originally published in Mediapost’s Search Insider February 25th 2010

The chart was the result of a neuro-scanning experiment we conducted with Simon Fraser last year. We were exploring how the brain responded to brands we like, brands we don’t like and brands we could care less about. The study was an ERP (Event Related Brain Potential) study. The idea of the study was to divide up the groups, based on their brand preferences and measure their brain waves with an EEG scanner. After, these waves were averaged and the averages of each group were compared with each other. What we were looking for were differences between the waves. We were looking for gaps.

It turned out we found two gaps. The brain waves are measured based on time, in millisecond increments. When we initially did the study, we were looking for something called the DM effect. This effect has been shown to represent a difference in how we encode memories and how effective we are in retrieving them later. We wanted to see if well liked brands showed different levels of brain activity when it came to memory encoding than neutral or disliked brands. The answer, as it turned out, was a qualified yes. What was most interesting, however, was the difference in the brain waves we saw when people were presented pictures of brands they love and brands they either dislike or feel ambivalent about. There was something going on here, and it was happening in two places. The first was happening very quickly, literally in the blink of an eye. We found our first gap right around 150 milliseconds – in just over 1/10th of a second.

The second gap was a little later, at about 450 milliseconds, or about half a second.

Brands = Faces?

Previous ERP work often used faces as the visual stimuli that subjects were presented with. Researchers like working with faces because the human brain is so well attuned to responding to faces. As a stimulus, they provide plenty of signal with little noise. What researchers found is that there were significant differences in how our brains processes well known faces and unknown faces. They also found differences in how we processed smiling faces and scowling faces. And the differences in processing showed up in two places, one in the 150-millisecond range and the second at about 450 milliseconds. We were seeing the same thing play out when we substituted familiar brands for familiar faces. So, what’s the big deal about about the 300 milliseconds that separate the two? Well, it’s the difference between gut instinct and rational thought. What we might have been seeing, as we stared at the projector screen, was two very different parts of the brain processing the same thought, with the first setting up the second.

The Quick Loop and the Slow Loop

Neurologists, including Joseph LeDoux and Antonio Damasio, have found that as we live our lives, our brains can respond to certain people, things and situations in two different ways.

The first is the quick and dirty loop. This expressway in our brain literally rips through the ancient, more primal part of our brain – what has popularly been called the Lizard brain (neurologists and psychologists hate this term, by the way). Why? Because if we hesitate in dangerous situations, we’re dead. So we have a hair trigger response mechanism that alerts us to danger in a blink of an eye. How quick is this response? Well, coincidentally, it’s usually measured in the 100 to 200-millisecond range.

But then there’s a slower loop that feeds the signal up to our prefrontal cortex, where there’s a more deliberate processing of the signal. If the signal turns out to be non threatening, the brain damps down the alarms and returns the brain to it’s pre-alert status. Cooler heads prevail, quite literally. The time for this more circuitous path? About half a second, give or take a few milliseconds. This more deliberate evaluation represents the second gap we saw in our averaged brain waves.

So…

Why was I fixated on that small gap between the squiggly lines at 150 milliseconds? It’s because this represented our immediate, visceral response to brands. Before the brain really kicks in at all, we are already passing judgment on brands. And this judgment will color everything that comes after it. It sets the stage for our subsequent brand evaluations. I have no idea what the 150 millisecond gap means. For that matter, I’m not sure what the 450 millisecond gap means. But I’m pretty sure it’s important. And I’m also sure that I’ll be spending a lot more time in 2010 digging into this difference.

Originally published in Mediapost’s Search Insider February 4th 2010

When making an organizational purchase, you have two types of personas involved: doers and buyers. Doers are the people who typically first recognize the need for your product. They will be the ones using your product. The question they are asking is “What does this product do?”. Their concern is not just what your product can do for their company, but for themselves personally. They need to know the features and benefits of your product.

The second persona who can be using your website during the B2B purchasing process are the buyers. These are the people who make the purchasing decisions and control the purse strings. They are concerned not only with the cost of your product, but also in eliminating the risk from the purchase. Buyers need to know who your company is, your company size and your financial stability. A buyer needs to know your company is established, dependable and reliable.

And this leads to your on-site content development strategy. Both doers and buyers need to alleviate risk and your website can help allay their concerns. Doers are looking for product information. Think of all the features and benefits your product can offer and if applicable, consider offering a free trial. You can build trust with doers by building your reputation as an industry thought leader through downloadable white papers or webinars. Buyers need to know you are a trustworthy vendor. They will be swayed by company information, references, case studies of success stories and testimonials.

The B2B buying process can be a long one, particularly the riskier and higher the cost of your product. But your website can be a key to removing friction points during this process – if you remember the types of people you need to serve up content to.

—-

An often discussed topic in the UX world is the eye tracking method and its right to exist as a meaningful and valid usability method.

I am including eye tracking in my usability tests as often as possible, because I found it to be an helpful method to investigate what user look at on a result page. Particularly the information on eye movement in combination with user behavior on the search results page provides a lot more information about the motivations for certain behaviors, than tracking user behavior alone. Eye tracking data should always be seen complementary to the live statistics. The same applies also to other forms of searches like maps, news or product.

Ian Everdell, Usability Consultant for Enquiro is working with me on advanced search issues for the t-online.de search since 2007. During the last month of our work we have frequent – sometimes controversial – discussions on theories and methods. To give you a chance to participate in the discussion we had, I summarized some parts here in my blog as an interview.

So, if you think “search usability” is cool stuff you should read on, otherwise I hope you will find something more interesting on my page.

SQ: The world of usability seems to divided up into two groups. On one hand we have the fanatics about eye tracking, and on the other hand we have the critics. So Ian, what do you think? Were do you see the benefits outweighing the costs?

IE: The eye tracking debate is certainly a heated one, especially when you have well-known usability figures like Jakob Nielsen and Jared Spool taking opposite sides. The argument is typically that eye tracking is a highly specialized and expensive technique that doesn’t add any insight that couldn’t have been found using more traditional methods. I respectfully disagree.

I think that eye tracking can add a lot of insight. It can useful for identifying areas of the page that seem to be receiving lots of or little attention, although it’s important to realize that gaze does not equal attention (you can be looking at something without paying attention to it) and vice versa (you can pay attention to something that you’re not looking at). It can also help identify typical scan patterns, which can help you position information and page components appropriately.

A good example of this would be testing the search function on a web page. Imagine that the search box is placed in the lower left-hand side of the page. Participants in the study find the search box and use it to navigate to their goal without any problem, and they all report that the experience of using the site was good. However, analysis of the eye tracking shows that almost all of them looked for the search box in the top right corner first – maybe you should move it there to help speed things up?

Eye tracking can also be used to prompt study participants to talk about their experience. By replaying the video of their session and letting them watch what their eyes were doing, it can help them remember what they were thinking or doing at a specific point. Then they can talk about their experience in more detail.

Another interesting thing that we’ve found in some of our work at Enquiro (Link) that examines gaze behaviour during search is that while in the end it takes most people a similar amount of time to choose a link, men and women process the page in remarkably different patterns – this is insight that would be tough to get from traditional usability testing.

However, eye tracking is ultimately just one tool. It needs to be used in combination with other usability testing methods to get the best results.

SQ: What do you think is the biggest benefit of eye tracking for search?

IE: As I mentioned above, eye tracking has already shown us a number of interesting things specifically related to search, like how people scan a typical SERP and that there are differences between men and women. It also shows us how long people engage with particular parts of the page, although we have to keep in mind that looking doesn’t equal attention, and that we don’t know whether extended viewing means that people are engaged or confused. However, knowing that something is happening in that area of the page allows us to ask the participant more questions about why they might have been looking at that part of the page.

One interesting finding that Jakob Nielsen has published using eye tracking data is that people really only read the first 11 characters of a headline – this has huge implications in how the title tags that are used for search results are written. Eye tracking has given us pretty definitive proof that using “Welcome to the web site for…” is not a good title tag.

SQ: Which usability methods do you find helpful when optimizing a search?

IE: Traditional usability testing is probably the best method – get people to interact with the page and then verbalize their thoughts, etc. when they’re done. This gives you a good understanding of what they were expecting (chances are that this day and age the answer will be “I was expecting it to be more like Google”). But I think that eye tracking is a good follow up to this because it can give you more understanding of where certain elements should be placed, how content should be written to promote reading and understanding, and ultimately coach a click through.

SQ: Which usability method do you find undervalued for search in the usability community?

IE: I’d love to see remote usability testing start to be incorporated into search – capture real users in their real environment doing real things. I’m sure the major search engines could do this fairly easily, but of course the volume of data would be enormous.

But then you have a different problem from traditional usability testing: instead of having to create intent for the user to get them to use a specific search term, you know the search term but not the intent. You can only infer it from the search term itself… and this is why behavioural targeting is tricky.

SQ: What are the typical mistakes that happen when testing a search in the usability lab? How should the structured questionnaire be developed? What types of questions are relevant?

IE: I think the biggest problem with testing search in a lab is that you have to craft intent for the participant (at least, if you want to compare between participants). There are some studies that allow participants to do “free searching”, where they are given a vague intent and are allowed to complete the task completely on their own – the problem with this is that each participant might use a slightly different search term, which leads to a different SERP, which leads to a different experience altogether, so it’s tough to compare between participants.

I think another inherent difficulty in testing search in a lab is that there are billions of possible experiences that people could have – one little change in the search term, the search algorithm, blended results (images, video, news, etc.), or a page’s metadata or content can change the SERP, which again makes it difficult to compare between experiences. So taking the few exposures that you get in a lab and generalize them to each and every SERP can be quite a stretch.

Another difficulty with testing search in a lab is that search is about more than just the SERP – it’s about what’s after the click as well. Users make judgments about the search engine based on what they find afterwards, so usability studies should include exposure to landing pages and web sites as well

Relevant questions for the usability of search include standard quantitative measures like time to click, a user satisfaction scale and eye tracking metrics (if collecting eye tracking). I would also be interested in subjective qualitative information from the participants like their perception of relevancy and their expectations for what would be on the page.

SQ: What are the right (or most valuable) metrics for measure the usability of a search? How do you choose them? Are they different for the different searches, e… local searches, web search, product searches etc.?

IE: The standard usability metrics are effectiveness (the user is able to accomplish the task, measured as a completion rate), efficiency (how quickly the user can finish the task, the number of clicks, or some similar measure of effort needed), and satisfaction (how happy the user is after performing the task).

However, I think the most valuable metric for search usability would be the correlation between relevancy (measured as part of satisfaction) and time to click (efficiency) – searching needs to be effortless for the user, so the ultimate goal would be to maximize relevancy and minimize the time it takes them to make a decision.

In terms of different metrics for different types of searching, certainly you might want to define what “effective” or “efficient” mean – effectiveness for a local search might mean finding a business on a map, while for a product search it might mean finding the price – but ultimately you want to focus on improving those three key metrics.

Choosing metrics to measure, for any type of usability testing, should be based on ensuring that the experience is as good as possible for the user – it has to be efficient, effective, and satisfactory.

Whether you choose to work on one of those at a time (e.g. make sure they can do the task, then work on making it faster/easier to do) or all at once, it all comes down to the fact that you’re there to serve your user.

Here at Enquiro we operate a little differently. Eye tracking is an ingrained part of our Usability studies, and we’ve built a strong reputation on our ability to get the best out of eye tracking technology. We don’t use eye tracking as a $30k upsell to lab based studies. It’s a critical component of our studies and just another tool in our Usability methodology. Whether its’ value is actually 5% or 30% of the total insight and value we create from a study is debatable. It’s certainly not the majority of the value, as that comes from our rounded insight into user behaviour, intent, and other contextual elements.  But we think that without the use of eye tracking, Usability practitioners are “unable to see what participants aren’t able to see”. There’s a degree of reliance on participant recall and self reporting, which has a certain level of inherent Hawthorne effect, social desirability, and central tendency.  In our mix we add eye tracking, and other hard metrics such as time to complete task, and task pass/fail, to create assessments.

So, if you’re being oversold on eye tracking, perhaps it’s the approach that’s wrong and not the methodology itself.