Author Archives: ghotchkiss

Another Year, Another SEMPO Survey

First published December 6, 2007 in Mediapost’s Search Insider

As chair of SEMPO, one of the things I’ve seen establish itself is the regular SEMPO state of the market survey. Every year about this time, we field a gargantuan survey (in excess of over 400 questions, including all the possible branches) in which we attempt to capture a relatively complete snapshot of where search is at this given point in time. We use the data captured to analyze year over year trends and attempt to project forward  how these trends might develop over the coming year. This year, in addition to the main survey, which primarily collects North American data, we’ll also be launching a smaller European version, scheduled for the new year.

One of the interesting things about the survey is that it can act as an early warning system for emerging trends. Among the biggest variances last year was the clear signal that many marketers would ideally like to move their search marketing management in-house. Over the course of the year, we’ve certainly seen this start to happen, and my suspicion is that we’ll continue to see this trend play out in 2008 and beyond. In-house marketers are the fastest growing SEMPO membership segment. By digging this point out of the data, we were able to put it on the radar of search agencies, letting them know that perhaps the value provided wasn’t quite up to what clients were looking for. It may not have been good news, but it was vital.

Two years ago, click fraud emerged as a major concern. While the urgency of click fraud seems to have lessened, it’s still a nagging doubt that seems to linger in the background and throw a shadow on the effectiveness of search as a marketing channel.

Other data points worth mentioning: Last year, for the first time, search seemed to pierce the consciousness of C-Level executives. This trend literally appeared out of nowhere, as one year previous these execs were barely aware of search. Heightened awareness on the executive floor might be responsible for some of the directives to bring search in-house. If you want something to capture the attention of executives, make it a budget line-item. For many companies, search broke this threshold in 2007.

Also, while many marketers seemed to feel there was still room to up their keyphrase bid amounts, the ceiling seemed closer than ever before. Pricing fatigue became a reality for more and more search marketers.

That was the backwards look at last year’s survey findings. This time around, it will be interesting to see how what has arguably been the biggest year of change in search has impacted marketers. With the rollout of unified (or universal, or blended, or 3D, or whatever the label du jour is) and personalized search, the interface has undergone some pretty dramatic changes. While the impact on most of our campaigns has been minimal to this point, we’re all thinking about how this might change the game. It will be interesting to see what changes in attitudes towards search as a channel boil out of the data.

So, if you have a few minutes to spare (about 10 to 15 should do the trick) make sure you take this year’s survey. Like any study, its accuracy is totally dependent on how many people volunteer their information. The more who take it, the more we can slice and dice the data and pull more nuggets out of it. And, by taking it, you get the first look at the findings.

Finally, a word of thanks to my SEMPO research co-hort, Kevin Lee from Did-It. He puts a lot of hours into question formulation and testing every year, all of them as a volunteer. This survey is really his baby.

What Makes a Rumor so Easy to Spread?

urban-legend-rumorWe all want to be part of the next viral world of mouth success story. We want our product to be at the epicenter of a “buzz” storm that spreads like wildfire across the internet. But the conditions that lead to true word of mouth viral outbreaks dictate that these outbreaks are few and far between.

Jumping the Weak Ties

First of all, let’s look at what’s required for word of mouth to spread. The trick to a true viral outbreak is finding something that will jump the “weak ties”. Mark Granovetter identified weak ties in a social network back in the 70’s. Basically, social networks are not uniform and even. They are “clumpy”. They have dense clusters, comprised of people who tend to spend a lot of time together. These are family members, co workers, close friends, members of the same church or organization. Word spreads quickly throughout these clusters, because of the frequency of communication and the nature of the relationships between the members of the cluster. There’s an inherent trust there and people talk to each other a lot. This makes the social ties within the cluster strong ties. Given this, once one person in the cluster knows something, there’s a pretty good bet that everyone in the cluster will know it in a relatively short period of time.

But the challenge comes in getting a message to make the jump from cluster to cluster. How does word of mouth spread from one group of co workers to a church group in another town? To do this, we’re relying on social ties that are much weaker than strong ties. We’re counting on an acquaintance to pass word along. And for that to happen, some conditions have to be met first.

Lowering the Drawbridge

In 1993, Jonathon Frenzen and Kent Nakamoto followed up on Granovetter’s earlier work (Frenzen, Nakamoto: “Structure, Cooperation and the Flow of Market Information,” The Journal of Consumer Research, December 1993) to see the conditions that had to be met before a message would jump across a weak tie. In their words,

“Instead of an array of islands interconnected by a network of fixed bridges, the islands are interconnected by a web of “drawbridges” that are metaphorically raised and lowered by transmitters depending on the moral hazards imposed by the information transmitted by word of mouth.

In their study, they looked at a number of factors, including the nature of the message itself, and the concept of moral hazard, or how it would impact the messenger. For the test, they used news about a sale. In one social network, they saw how fast word would spread about a 20% off sale. In the other social network, they used a sale where the discounts were a more remarkable 50 to 70% off. To introduce a moral hazard variable, they also altered the availability of sales items. In one case, quantities were very limited, and in the other, quantities were practically unlimited.

What they found was that amongst strong ties, word of the sales spread fairly quickly in most instances. But when the message wasn’t that remarkable (the 20% off example), word of mouth had difficulty jumping across weak ties. Also, when moral hazard was high (quantities were limited) again, the message tended to get stuck within a cluster and not be transmitted across the weak ties.

Mexican Vacation Sale

Let’s use an example to make this a little clearer. Let’s imagine an airline is having a seat sale to Mexico. In the first example, it’s $50 off per seat, but it applies to every seat on the plane, on every flight. There is no limit on the inventory available. In the second instance, instead of $50 off per seat, the entire cost of a return flight to Mexico is just $50. That’s much more remarkable. And in the third instance, the sale is again $50 per person, but it’s limited to 10 seats on 2 flights, for one day only. Only 20 tickets are available at this price.

In the first instance, you would probably only pass along the information if someone happened to mention to you that they were thinking of going to Mexico. The information is not that note worthy. The value of information is not that great. There’s little chance that this would ever move beyond your “strong tie” cluster. It’s not something you’d go out of your way to mention to an acquaintance.

In the second instance, a $50 flight to Mexico is big news. And we’re socially predisposed to share remarkable stories. We believe it elevates our social status within our cluster. Every one likes to be the first to tell someone about something remarkable. It’s part of human nature. So we’ll go out of our way to share this information. We don’t even wait for someone to raise the topic. This is noteworthy enough that it merits bringing up in any context. It’s worth interrupting normal conversations for. Word will spread far and wide, across strong ties and weak ties alike.

But in the third instance, even though the news is remarkable, we personally have something to lose by spreading the story. There are only 20 seats available, so if we tell too many people, we might not get a chance to take advantage of the sale ourselves. Chances are, we won’t tell anyone until our seats are booked. And even then, we’ll probably only tell those we’re closest to. After we look after ourselves, our next inclination is to make sure those that are closest to us won’t miss out on the opportunity. Again, because of this “moral hazard” there’s little likelihood that word will spread beyond our strong ties.

Rumor has it

So, now that we know the limitations of message transmission within a network, depending both on the structure of the network and the cooperativeness of it, let’s look at one type of information that always seems to spread like wildfire through any social network, regardless of the circumstance: the juicy rumor.

Rumors have no moral hazard, at least, not for us. There are no limitations of quantity. We don’t stand to lose out (at least, not in a material sense. We’ll leave the ethical questions aside for now) by spreading a rumor. So that restriction is gone.
Secondly, the likelihood to spread a rumor depends on the nature of the rumor itself. First of all, does it involve people we know? Personal rumors about people we know are almost irresistible to spread. They beg to be passed on, again, because they put us in the position of “being in the know” and having access to information not available to everyone. Second to the personal rumor is the celebrity rumor. These are a little less “spreadable” because we’re not in the same privileged informant position. Also, although we know the people involved, in the public sense, we don’t really know them in the personal sense. When it comes to rumors, the closer to home they hit, the better.

Finally, we have the “juiciness” of the rumor. How sensational is the story? How remarkable is it? A rumor about your neighbor’s washing machine breaking down isn’t going to go too far. But an affair leading to a marriage break up, being fired from a job or a significant health issue, unfortunately, are stories made to spread. Because we’re human and inherently competitive, we love to spread bad news about others.

Fine Tuning the Rumor

And this brings us to an almost universal behavior seen whenever rumors tend to spread. We like to fine tune the story to make it a little more interesting. Rumors are subjected to “flattening”, “sharpening” and “assimilation”, just to make the story a little more sticky. Flattening is where we get rid of the details that get in the way of what we feel is the noteworthy aspects of the story. In some cases, the discarded details are contradictory and in some cases they’re just extraneous. Regardless, if they’re not pertinent to the main story we want to get across, or if they dilute the story, we toss them out.

Sharpening takes the remaining facts and enhances them a little (or a lot) to bring the story and it’s value as news into sharper focus.

Finally, assimilation is where we take the story and make sure it fits within our shared mental framework. We alter the story so it fits with ours (and our recipients) shared beliefs and views of the world. That’s one reason why rumors are so “spreadable”. We alter the story to ensure it’s interesting, and the further the story goes, the more irresistible it becomes.

The ultimate example of this are urban legends, where once there may have been a kernel of truth, but the stories have become so flattened, sharpened and assimilated through countless retellings that now, as intriguing as they are, they are basically manufactured fictions.

Negative Word of Mouth

We’ve always known that negative word of mouth spreads faster than positive. When we take what we now know about social networking and apply it, we begin to see why. For instance, negative word of mouth and rumors share a lot in common. There’s generally no moral hazard in play. In fact, the reverse is true. You’re actually helping people out by sharing this information, and you get a little retribution and revenge yourself. It’s a twisted win-win!

And for some reason, humans are much more likely to pass along negative information than positive. Again, it comes to our concept of social hierarchy and building ourselves up through the misfortunes of others. Admirable it’s not, but predictable? You bet!

And finally, the better known a company or brand is, the more likely negative word of mouth will spread. If there’s bad buzz circling about Nike, McDonald’s or Starbucks, we’ll all take part because all those brands are part of our shared frame of reference. We’ve already assimilated them.

By the way, remember that negative word of mouth will also be subjected to flattening and sharpening, as well as assimilation. So the negative buzz will get worse with each retelling.

Obviously, if you’re counting on word of mouth as your marketing channel, you have to take the reasons why word of mouth spreads into account. It can be made to work for you, if the conditions are right, but remember, this is not a process you have much control over. You can plant the seeds, but then human nature will take it’s course.

The Whys of Buy: Visualizing the Buy

Human brainVisualization risked becoming another one of those clichéd words through the 90’s, because it was used by every self improvement guru as a path to success. Visual success and it will be yours. But the fundamental principles of visualization bear up, in some very practical and surprising examples. And the neurological science behind visualization is sound.

Visualization allows us to sense a scene in our minds when we read a passage. Actually, the word visualization is a little misleading, because it only refers to the sense of sight. Visualizations can engage all the senses. For example, we took our two daughters to Manhattan last summer, landed at 11 pm, and because we weren’t tired, walked through Times Square at midnight. New York was in the middle of a heat wave and the temperature was still 98. The combination of heat and humidity added a particular edge to the smell of garbage in the streets, that sickly sweet/sour odor that punctuates the more appetizing smells wafting from restaurants and street vendors. Times Square was still going at full tilt (this was a Saturday) so the din of taxi horns was deafening. At every corner, we still had to elbow past street vendors and crowds jaywalking through the intersections. For my daughters, it was a rude sensory awakening to the Big Apple.

As I was writing that, feelings, sights, sounds and smells were being activated in my mind. I was recalling the images, and could, in my mind, feel the humid heat, smell the odors, hear the horns and see the crowds. If any of you reading this had been to Times Square on a hot summer night, you probably have your own scenes, from your own experiences, replaying in your mind. But the amazing thing is, if I say falafel stand, you can see, smell and perhaps even hear it. That’s because the same parts of your brain are firing that would actually be activated if you were physically there. Imagination is the next best thing to being there.

Athletes have long known this. Visualization starts building the same neural pathways that actual physical action does. A golfer struggling with his swing can visualize it and improve it, without a club in his hands, because he’s giving his brain a trial run. The same is true with a gymnast learning a new move. Studies have shown that imagining a 5 finger piano exercise results in a significant improvement in performance.

But perhaps the most startling evidence comes from a study done by the Cleveland Clinic Foundation in Ohio. Here, from About.com, is the summary of the study:

They split 30 healthy young adults into 3 groups.

For 15 minutes a day, five days a week for 12 week, Group #1 imagined exercising their little finger muscle. Group #2 imagined exercising their biceps muscle and Group #3 acted as a control group and did no imaginary exercise. Those in the first two groups were asked to think as strongly as they could about moving the muscle being tested, to make the imaginary movement as real as possible. The researchers measured muscle strength before, during and after the training sessions.

Group #1 (the finger exercisers) increased their strength 53 percent, and Group #2 (the biceps group) increased strength by 13.4 percent.

These results are somewhat unbelievable. Simply imagining exercise can make you stronger! Literally without lifting a finger. That’s the power of visualization.

So what does this mean for marketing? Visualization plays a part here as well. We often visualize our way through a purchase. If we’re looking at buying a car, we visualize ourselves driving it. If my wife is determined to buy a dress, she visualizes herself wearing it. Even if you are suddenly craving something from Starbucks, you can see, smell and taste the coffee before you ever get it in your hands. Visualization is a powerful part of purchasing, and once we build these neural pathways, it takes us much closer to the actual purchase. Smart marketers start building the pathways before you ever set foot in the store. That’s why personalized products can be so powerful. Personalization forces visualization.

Of course, visualization of product usage is nothing cutting edge. Most marketers do this instinctively. But what about visualization of the actual purchase itself? How can you start building the neural pathway required to ensure the transaction is completed? This is particularly important in more involved purchases, such as trips, cars, houses or more involved B2B purchases. In each of these cases, the very act of buying can act as an obstacle to a sale. It requires time, commitment and knowledge. For all these reasons, a little mental practice could improve the odds for success. Let me share another example.

In the 1960’s, social psychologist Howard Leventhal wanted to persuade a group of college seniors at Yale to get a tetanus shot. What he wanted to test was whether fear would be a more powerful influencer. So several information booklets were produced. Some were “high fear” with graphic pictures and descriptions. Some were “low fear”, with a more toned down, informational approach. The booklets were distributed and, somewhat predictably, the high fear booklets seemed to be more persuasive. The groups that received these booklets were more convinced about the importance of shots and more of them indicated that they intended to get inoculated. But one month later, almost none of the participants from any of the groups, high fear or low fear, had actually gone for an inoculation. A mere 3 percent had actually been inoculated. This was an unforeseen glitch in the experiment.

So Leventhal redid the experiment, but this time with one small change. This time, in all the booklets, he included a map showing where the clinic was and the hours it was open. This time, the inoculation rate went up to 28%.

If we look at the power of visualization, the thing that surprised Leventhal really isn’t that surprising at all. The first round of the experiment did a good job of inducing the visualization of consequences, in this case, negative consequences. The high fear booklet let the students visualize what might happen if they didn’t get a tetanus shot, and so it was persuasive. But it didn’t close the loop. It wasn’t that the message wasn’t persuasive. It was just that it left the door open for life to get in the way.

But the second version allowed the student to visualize the path required to actually get the inoculation. I’m sure most of them probably knew where the clinic was, but the inclusion of the map prompted them to visualize actually going there, and the hours allowed them to visualize where in their schedule they could fit in the visit. Once the students went through the mental process of visualizing action, there was a much higher probability that the action would take place.

What are the takeaways? If your purchase process requires a commitment on the part of the buyer, let them visualize the path required to get to the end. Use your website to build the path required to navigate through things like financing, negotiation, customer service, delivery and selection of products and options. Don’t just stop at visualization of ownership. Think about the visualization of the act of buying as well.

Passing the Tactical Torch to the New Kids

First published November 29, 2007 in Mediapost’s Search Insider

After the recent SMX show in London, I had a chance to have a wrap-up dinner with conference chair and Search Engine Land Executive Editor Chris Sherman. Chris and I, both feeling a little long in tooth, realized that there’s been a generational transition in search. The new generation is taking over the tactical stuff. As Chris said, “This blows away the traditional block and tackling stuff we used to do.” These are hotshots that live and breathe social media optimization, get a visceral rush out of an elegant link baiting campaign and measure their prowess through the number of Diggs they collect. They’ve taken organic optimization to a new level.

The Slow Surrender of the Sluggish Synapse

It was somewhat ironic, as I spent my sessions at SMX talking about things like bounded rationality, working memory and satisficing. To me, the working of the human mind is infinitely fascinating and that’s where I’ve been spending my free hours. I’m quite content to leave it to the up-and-comers to scramble up the listing hierarchy to grab the top slots. I’m more interested in what happens from the user side when they’re presented with those listings. Of course, I have the luxury of having a talented team working with me that can focus on the tactics while I play in my little strategic sandbox.

It reminded me of a passage I remember reading somewhere. A mathematician’s washed up by the age of 35 (I know, this is a point of controversy and I’ve read arguments on both sides, but I’m just using it as an example, so don’t get all worked up about it) but philosophers only hit their stride well into middle age. There’s a difference between sheer mental acuity and wisdom. Now that my synapses are slowing down, wisdom is really the only option open to me, so I’m grasping at it with both hands.

Wisdom: The Consolation Prize for Growing Older

I think it’s generally true that younger people tend to flex their mental muscles by solving puzzles of defined scope. They concentrate on the question ahead of them and revel in pushing the limits, punching holes in traditional thought and taking on risk that would prove unpalatable to a more pragmatic middle-ager, all in search of a solution that allows you to say, undeniably, that you’ve won. There are definitely winners and losers in the game of SEO. Number one is a winner. Everybody else is a loser, although in this case, the degrees of losing increase as you move down the page. It’s like sports. Nobody remembers second place. This appeals to the bravado of youth. SEO is for the young, and the young at heart.

But the question of who wins is a little more difficult to determine if you’re asking “why” questions. Why do people do what they do on search engines? Why do they make the decisions they do? Why do they pick certain brands over others?

I think unraveling the answers to “why” questions require patience and a more seasoned world view. There are fewer “aha” moments that signal victory. Answers are teased out little by little and added to the general body of knowledge about why we do what we do. The qualities that lend themselves to this approach come with age. They require being students of human nature. I’ve found that as I’ve grown older, I’ve become less frustrated with human frailty and more fascinated by human complexity. Of course, I’ve also become crankier. All of which makes me difficult to live with.

Search’s Big Picture: Step Back and Refocus

This trend also speaks to a maturity in the search space. It’s encouraging to know that search has started to develop a “big picture” that allows for strategic thought. Search was exclusively tactical in its early days, because its limited, siloed scope made it so. But now, search has become so integral in so many activities, we find overlap in almost everything we do. I can find much common ground in how we make decisions and how we use search engines. The top of this particular box is starting to open. And the broadening of approaches to optimizing search both as a marketing channel and as a human activity is healthy. As author Daniel Pink said, we need to develop our right-brain skills, “such as forging relationships rather than executing transactions, tackling novel challenges instead of solving routine problems, and synthesizing the big picture rather than analyzing a single component.”

So, as we walked down Edgware Street in London looking for a restaurant (we ended up finding quite a good Lebanese place), Chris and I talked about this passing of the torch and came to the conclusion that we’re okay with it. To be honest, I really don’t have a lot of interest any more in the tactics of search. That doesn’t diminish the importance of them; it just means that I’d rather do something else with my time. “What” doesn’t really capture my attention anymore. But “why”? Now there’s a question I can sink my teeth into. At least, while I still have teeth.

Brand Live and Die Face to Face

iStock_000004520845XSmallThe more I dig, the more I’m convinced that a big part of a brand’s success is the quality of its customer touch points, specifically, the face to face ones. Consider this overwhelming evidence:

The more emotion there is in an experience, the more vividly we remember it. It’s known as imprinting. So if we have very positive or very negative experiences, we remember them longer and more completely. Let’s say we visit a restaurant. If we have a terrible experience, we’ll remember it forever. If it was an amazing experience, again, we’ll remember it forever. If it’s mediocre and falls in the middle, it will tend to fade away.

Our memories are altered by the context in which we remember them. Let’s go back to our restaurant example. Whatever our experience, we will tend to alter it if we’re talking to a person who also had an experience with the same restaurant. If they had a great experience, but ours was negative, we’ll tend to alter our memory to make it more positive. Alternatively, if we had a positive experience, but someone else’s was terrible, suddenly we’ll alter our memory to make it less positive. This doesn’t tend to swing memories all the way from good to bad, but it alters and reshapes memories to better fit the context of recall. And over time, it can erode a once very good memory, or build up a rather negative one. Memory is not an accurate snapshot of an event, it’s a malleable story. So consistency of experience is important.

We get a much richer channel of communication when we’re face to face with a person. Studies have shown that receive only 7% of our communication from the words that are used. The other 93% is a combination of body language and tone of voice. So no matter how carefully you script your frontline customer encounters, the success will depend on the person delivering the message. We have very finely attuned credibility detectors.

The quality of the face to face interaction is the biggest factor in how satisfied we are in a product experience. Malcolm Gladwell used the example of doctors being sued for malpractice.

“Believe it or not, the risk of being sued for malpractice has very little to do with how many mistakes a doctor makes…. Patients don’t file lawsuits because they’ve been harmed by shoddy medical care. Patients file lawsuits because they’ve been harmed by shoddy medical care and something else happens to them.

“What is that something else? It’s how they were treated, on a personal level, by their doctor. What comes up again and again in malpractice cases is that patients say they were rushed or ignored or treated poorly. ‘People just don’t sue doctors they like,’ is how Alice Burkin, a leading medical malpractice lawyer, puts it. ‘In all the years I’ve been in business, I’ve never had a potential client walk in and say, “I really like this doctor, and I feel terrible about doing it, but I want to sue him.”

Medical researcher Wendy Levinson found that doctors that weren’t sued spent 3 minutes more with patients than those that were (18.3 minutes versus 15). But it wasn’t just time, it was the quality of time. More simply, it was the tone of the doctor’s voice. Recordings of interactions with doctors were recorded and then were played back for study participants, who then put the doctors into two groups, those that would be sued and those that wouldn’t be. The recordings were altered so participants couldn’t hear what was said, all they could judge was the tone of the voice. And even with this, they were able to judge with amazing accuracy which doctors would be sued. It wasn’t what was said, it was how it was said.

When you look at corporate examples, the power of person to person connections are clear in cases like JetBlue and Saturn. In both cases, the extraordinarily high level of customer satisfaction was due primarily to the quality of the face to face encounters. JD Powers rated the Saturn among the highest vehicles in terms of satisfaction not because it was a better car. It was because their dealer network didn’t follow the typical industry model, which was more like a school of piranhas. JetBlue’s employees had a mandate: make flying coach suck less.

Why is this important to remember? Because of the coming workforce crisis. The baby boom is shifting the majority of our workforce to the end of their working lives, and there’s a severe shortage at the entry level, typically the recruitment bed for service based businesses. This means good people are going to get tougher and tougher to find.

Also, there’s a move to cut costs by streamlining and outsourcing those vital customer touch points. Self serve customer service models are becoming more common, and in many cases, they’re backed up by a customer help line that’s been outsourced to an overseas call center. The call center has been provided the appropriate scripts, and, in most cases, adequate training on how to field a complaint. But, as we’ve seen, that’s really only 7% of the problem. The other 93% is connecting with a person who really cares about your problem and is trying to help you. That’s something you can’t script.

Let me give you an example. My wife and I recently flew to Lisbon on British Airways. We had to connect through Heathrow. I booked my flight directly through BA, but my wife flew on points, so that flight was booked through a partner airline. Both flights had less than an hour layover in Heathrow, and we had to change terminals. I didn’t really notice this at the time of booking, but soon, my partner airline notified us that they had moved my wife back to a later flight to allow her to make the connection. As anyone who has connected through Heathrow will tell you, the odds of making a connection with less than one hour is slim to nil.

I called British Airways to get my flight pushed back and was connected to what was obviously an overseas call center. The person on the other end, if they were considering a medical career, would be a sure bet to be nailed with a malpractice suit. The manner was brusque and indifferent. He informed me that they could change the flight, but there would be a $200 change fee, about 1/3 of the total cost of the flight. Plus, I would have to pay any difference in fares. I tried to explain to the person that the layover time wasn’t adequate and that BA screwed up with the initial booking, but to no avail. Finally, I hung up in frustration, to allow myself to cool down a little.

I resigned myself to the fact that I was going to have to cough up the extra $200, and phoned back a week later to make the change. This time, I got a much friendlier person who looked up my reservation and informed me that my flight had automatically been pushed back because an hour wasn’t an adequate connection time. I asked when this had happened and what had triggered the change. They said it was a flag that was automatically put up in the system so many days prior to a flight and had nothing to do with my previous call. It was the system correcting itself.

Everything worked out okay with BA, and the flight was actually one of the best transatlantic flights I had. But the poor quality of one encounter left an overall negative impression rather than a positive one. And, as reinforcement of it, when I was talking to a friend who had recently flown to Spain on British Airways, they had had exactly the same problem. Our respective memory retrievals quickly turned into a BA-bashing spree.

Realize the importance of person to person, and if you have to short cut anywhere, don’t short cut here. It’s the most important part of your business.

Yahoo! Big Idea Chair

Yahoo BICAn hour of make up and that’s the best you can look?

Yes, that’s the best I can look. The Yahoo! Big Idea Chair Shoot in Toronto is wrapped and the results are in. If you’re interested, you can see the ad pdf and the video they shot.

http://ca.advertising.yahoo.com/BIC.html

Thanks to Maor and the Yahoo! Canada team for thinking of me. Despite my disparaging remarks about makeup, it was a lot of fun and an honour (Canadian spelling intentional) to be considered.

It was also pretty exciting to be considered along with Nancy Vonk and Janet Krestin, the brilliant ladies behind Ogilvy’s famous Dove Evolution campaign (which cleaned up at Cannes), Laura Gaggi and my friend, Mitch Joel of Twist Image. Nancy and Janet, I often blow off about agencies not getting it, but your work was simply breathtaking on this one.

And yes, Bewitched actually did get me into advertising

The Whys of Buy: Impulse Buying

chickenebayThe other day, we were talking about what makes us buy (an appropriate topic for today, the biggest shopping day in the US) and Barb Newman, our General Manager, wondered what made us impulse buy? She was trying to figure out why she had dropped way more money than she intended on a purse. Being intrigued by the buying mechanism that seems to be locked in our skulls, I decided to do a little digging to find out what’s going on when we just seem to pick up something off the shelf on sheer whim.

Spool’s Impulse Buying Study

On doing some digging, I found a study done by Jared Spool, a usability consultant I have a tremendous amount of respect for. And Jared found that market research, as I posted about a few days ago, has its limits. As Jared’s starting point, he looked at survey’s conducted by The Yankee Group and Ernst and Young. Both surveys asked why respondents would make impulse purchases on the web. With the Yankee Group survey, 75% indicated because of price. Ernst and Young’s survey said that 88% of purchases were made due to price. Again, these were surveys where buyers were asked to rationalize their behavior. And saving money seems like a pretty rational reason.

But Jared wanted to see what real people shopping for real products did. As most usability people do, he wanted to observe real world behavior. So his group got 30 people, who had real things they wanted to buy, gave them some money and sent them on an online shopping trip to a few preselected sites that had what they were looking for.

What they found was significantly different that what the Yankee Group and Ernst and Young surveys showed. While many of the participants bought what they were looking for, a significant number, 34%, also added other items into their shopping cart that weren’t on their original lists. Was it because the prices were irrestible? No, in fact, only 8% of the impulse purchases were because of price.

Jared and his group purposely picked out shopping sites that had promotional offers and seasonal sales in prominent display positions, especially on the home pages. But very few of the purchases were of these sale items. The impulse buys were spread across 41% of the sites in the study, including everything from pet shops to computer accessory stores. Almost none of the items were on sale. They were just things that suddenly tweaked the shopper’s interest.

Here’s the other interesting thing about the study. Most of the impulse items were chosen while browsing through the category pages. They had chosen a category based on what they were shopping for and had found related items that struck their interest and were subsequently added to their cart.

The Nucleus Accumbens Made Me Do It

So, why do we impulse buy? I’m still not sure, but here are some hunches, based on some of the other research I’m doing in the mental mechanics of buying.

A study earlier this year by Carnegie Mellon, Stanford and the MIT Sloan School of Management might be able to shed a little light on Spool’s findings. Using fMRI imaging, they also gave participants money to go shopping. They then monitored activity in various parts of the brain.

They found that when we anticipate buying something, the pleasure center, the nucleus accumbens, is activated. We begin picturing ourselves in possession of a product and visualizing ourselves using it. We start to build neural pathways that reinforce what it would be like to have the product. But, if the price is excessive, the study found that the brain has a shut off mechanism.  A part of the brain known as the insula is activated and the part of the brain we use to balance gains versus losses, in other words, is the product worth the price, the medial prefrontal cortex, begins shutting down. We literally put the purchase out of our mind because the price is more than we’re willing to pay.

So, let’s go back to Jared Spool’s study. I suspect we get into shopping “modes” where the parts of the brain associated with acquisition of a product sustain some activity. We’re prepared to buy, so the nucleus accumbens kicks into gear and keeps firing. We’re in “buy” mode. And we’ve accepted that we have budget available. We start out looking for the product we intended to buy, but, on the way, if we see something we also decide we need, especially in a related category, our “buying” mechanism is already activated. We’re already primed to consider purchase. We’re not looking for a bargain (although finding one certainly wouldn’t hurt), but by the same token, an outrageous price would probably shut down the process by kicking in the insula. Think of the insula as the brain’s sprinkler system, snuffing out any impulsive sparks before we burn ourselves. As long as the price is reasonable, and doesn’t introduce significant “pain” we’re more likely to purchase.

The fMRI study also showed that once we flip into buy mode, we tend to stay in this groove. This is why it’s much more dangerous to shop with credit cards than cash. Credit cards allow us to put off the “pain” that might kick in the insula, letting the nucleus accumbens have its way. When cash runs out, it runs out. It forces us to pay more attention to the “pain”.

In Spool’s study, the pain had been effectively removed by giving the participants money to spend. And by browsing through categories where they already had interest, there was a greater likelihood to pick related products and purchase them through impulse. Bargain basement prices really had nothing to do with the process. It’s just that most of us don’t understand the mechanics of buying that happen at the subconscious level.

Back to Barb’s Purse

So let’s get back to Barb’s purse. Was it really a impulse buy? Well, not really. As I chatted more with Barb, she indicated that she had seen the purse earlier in a magazine, fallen in love with it, but the price was much higher than she wanted to pay. So Barb’s nucleus accumbens had gone into overdrive, but Barb, being a practical shopper, had quickly doused the flames when her insula kicked in. The pain was too great to make a purchase.

But, a few months later, she’s in the mall and sees that the store that carries the purse was having a 25% off everything in the store appreciation sale. Suddenly, the nucleus accumbens is reactivated, primed by all the visualization that Barb had done since first seeing the purse, thinking how great it would be to own it. The 25% off sale lowered the pain threshold enough to keep the insula from kicking in, and the next thing Barb knew, she had put down a deposit and put the purse on layaway. She didn’t know what hit her. Now, she knows it was a little bit of gray matter hiding deep in subcortal brain called the nucleus accumbens that’s to blame. But this wasn’t a true impulse buy. It was more like a delayed buy.

So, if you overspend today, remember, it was the nucleus accumbens that made you do it. Try explaining that one to your significant other.

Why We Have to Keep Doing Market Research

Following up on my previous post about the problems with most market research, here’s a plea why we should keep trying to get it right.

At the recent London SMX show, I presented on the Ad Testing and Research panel. Like other times I’ve done this panel (this is probably the 3rd or 4th time) I hear about skillful practitioners employing various A/B and multivariate testing methodologies. Ad testing is a definite must do, but before my presentation, which came at the end of the session, I took a few minutes to provide an alternative point of view.

I asked the small crowd how many of them were doing regular campaign management, checking click through rates, conversion rates and optimizing their campaigns based on what they saw. Almost everyone put up their hand. Then I asked how many did A/B testing. This time, a little more than half put up their hands. Next, I asked how many were doing multivariate testing. This time, about one third of the crowd. Finally, I asked how many had actually sat, watched a customer interact with their site and then asked them questions. We dropped down to about 10% of the group, and most of these were in a fairly structured usability test, with limited or no opportunity for interaction with the user.

Now, campaign optimization, A/B and multivariate testing are all best practices and should be done religiously. But I urged the marketers in the room to step back from their data heavy, spreadsheet  bound view of the world and pick up a book on cognitive psychology, social science or simple usability. Better yet, spend some time just watching how real people interact with your site. Try, for a moment, to look at the world through your customer’s eyes.

The problem with the typical, quantitative methods are that they’re all lagging indicators. You don’t get an idea of what’s happening until after customers have interacted with your ads and your site. You generally get a good sense of what they did, but it’s very difficult to determine why they did it. To do that, you have to dig beyond the numbers. You have to try to get into that subconscious mind. And that’s not easy. Typical market research methodologies won’t cut it. To get some idea of what’s required, read Clotaire Rapaille’s The Culture Code, or Gerald Zaltman’s How Customer’s Think. Do some digging into the work of Herbert Simon.  It takes a deft combination of psychiatric know how and detective skills. But here’s why it’s worth it.

For the past Century, we’ve largely refined our marketing practices based on trial and error. Pretty much everything has been done through seeing what’s worked, changing something, and seeing if it worked better. That’s been okay, as long as the channels we used to reach customer’s were relatively limited. With limited channels and a certain amount of control inherent in the process, we could do this. But those days are over.

Now, rather than a few controlled channels that run pretty much straight from the advertiser to the customer, we have an explosion of information that turns the typical buying process into a Gordian knot of unbelievable complexity. We can’t control the variables anymore. When there are so many channels, so many interdependent factors and so much of it affects customers below the conscious level, trial and error is just not an effective testing methodology anymore. In fact, it was never an effective methodology, for all the reasons I stated in my previous post. It’s just the best we had.

Let me use another example. The way we did marketing was pretty much like jumping in a car, randomly making decisions whether to turn right or left, keeping track of our success rate in getting nearer to our destination, and using this method to eventually pick the right route. This method might eventually work okay in a town of a few thousand people, but try doing that to navigate through New York or Los Angeles. We don’t have enough time in our lives to leave this much to chance. A map (or better yet, a GPS) is a much better alternative.

But we’re just starting to put that map together. And it won’t come from market research. Market research, at least in it’s current incarnation, is hopelessly flawed. It will come from diving deep into the workings of our brains. And once we begin putting the map together, it will allow us to begin to measure leading indicators. It will keep us from the trap of relying on self reported rationalizations and dig deeper into all the activity that’s happening below the conscious surface of our minds. That’s where the answers will be found.

Here’s another reason. Our brains are not only complex, but they’re also highly adaptive. As we do new mental activities more often, and abandon previous ones, new routes are established through the neurons and old ones become overgrown and eventually, unused neurons are cut away. It’s called “pruning” and “neuroplasticity”. It’s probably why you’re much better at using a search engine now than you are doing the geometry you learned in grade 9. We’ve worn new paths in our brain.

This is also true of how we’re buying. The way we buy now is bearing little resemblance to the way we bought in 1975. As time goes on and we rely on the Internet more and more, the paths that we used to use for our consumer decisions will become overgrown and we’ll clear new ones. This will happen not only at the conscious level, but also the sub conscious level. We will literally rewire how our brains decide what to buy. So the body of market research that has laboriously been gathered over the past several decades will become obsolete. And to discover those again through trial and error will be an long and potentially impossibly task.

So, a word of advice. Step back from the spread sheet now and again. Take a break from looking at “what” and start to explore “why”. Dig into things like the triune brain, selective perception, bounded rationality, working memory and some other basic cognitive concepts. It will be time well spent.

What’s Wrong with Market Research

sharingbrainWhen we first started doing research at Enquiro into how people used search, we found very quickly that what people say and what people do are very different things. It just happened that we were doing a survey and a focus group at roughly the same time. In the survey, where we got the results first, we asked if things like the position of a listing was important in whether people read it or not. We asked people to rank a number of factors on their relative importance, including position, relevancy and trust in brands and vendors shown. Almost without exception, in the survey, people indicated that relevancy was the key factor. They also indicated that they read listings pretty carefully and gave a fair amount of thought before selecting one. Finally, many said they would never click on a paid listing.

Then, we invited about 30 people into our labs and actually recorded their interactions with the search engines (before our eye tracking studies) and it quickly became obvious that how they said they used a search engine and how they actually did were two different things. The vast majority of clicks happened in the first few listings. Many who indicated they wouldn’t click on paid listings actually did, and perhaps, most interestingly, the average interaction was around 10 seconds or so. Subsequently, we’ve seen this type of behavior repeated in eye tracking after eye tracking study. Of course, the famous golden triangle study we did with Eyetools and Did It, and subsequent ones conducted by Enquiro, have shown over and over how quickly we interact with a search engine and how much of our scanning activity is “top loaded”. Also, we don’t really skip over sponsored listings, but in some circumstances (research based activity) we’re less likely to click on them. We’ve used this body of research to come up with a fairly consistent model of how people interact with search results. The results belie what people indicated in our very first survey. Well over 60% of the clicks happened in the first 4 or 5 listings, including the top sponsored ones. People generally spent just a few seconds on the page (around 10 to 12 seems to be the average) in which they scan (not read) 4 to 5 listings. There was almost no deliberation. People click quickly, and if they don’t like what they see, they click back. It would take the average person about 2 minutes to actually read all the results on the average search results page. Even if we just read the top 4 or 5, we’d be spending about 30 to 40 seconds on the page. It takes about 7 seconds to read one listing. But we don’t spend much longer than this covering 4 to 5 listings, about 2 seconds per listing. Obviously, we don’t give a lot of thought to the credibility of the search listings.

So, were all 1600 of our original survey respondents liars? Were they intentionally misleading us? No, they were just being human.

What we found was the systemic fault with almost all market research. And there’s a very good explanation for it. We’re generally not aware of 95% of what we do or why we do it. That’s because much or what we do is hidden in our subconscious. I’m currently reading How Customers Think by Gerald Zaltman and he pinpoints the problem with traditional market research. In almost every case, we ask people to tell us, either verbally or through writing, what they’re thinking. Just by doing this, we kick in the cortex, the rational seat of our intellect. But Zaltman tells us that at least 95% of every decision is made subconsciously. There, in the murky depths of our brains, predating the evolution of our cortex by many millions of years, thoughts are created through tremendously complex connections of memories, beliefs, instincts and intuition. In many cases, our decisions are made long before they bubble up to our conscious minds. The conscious mind exists to put a little polish on them and, in most cases, to rationalize a decision that was largely based on primal instincts. We may have done what we did because our flight or fight mechanism kicked in, or because our need to procreate surfaced. That’s why we chose the minivan, or the red convertible. It really had nothing to do with the Consumer Reports rating. But, being highly evolved humans, we convince ourselves that our choices are much more rational than those of a lizard (our basic brain core, which rules many of our decisions, is basically the same as a reptile’s brain).

In our case, our initial respondents indicated that they deliberated over which search result they chose. In actual fact, there was little risk in choosing a wrong link (it’s not like our lives, our family or our money is at stake), so we cut off the amount of deliberation we did and after a quick scan, picked the result that seemed to be most relevant to our intent. The lack of deliberation wasn’t lack of intelligence, it was a survival instinct bred into us by eons of evolutionary refinement. If there’s no immediate risk to us, why should we kick in our brains and spend unnecessary time and cortex processing power to come to the optimal decision. It’s not required. A simple scan and click will suffice. Our brains are simply doing what they’ve been programmed to do. And it’s not that the decisions are bad. As Malcolm Gladwell shows in Blink, often these decisions prove to be better than the ones that we endlessly deliberate over. Our brains, especially the 95% that remains under the surface, are amazingly adept at making good decisions.

But there’s a more fundamental issue here. If what we experienced in search is typical in all market research (which it is) how do we ever find out how people actually make purchase decisions?

This is a significant challenge, the extent of which might not be obvious at first glance. Let me use an analogy to further illustrate. Remember the tale of the shoemaker and the elves? Let me use that and adapt it slightly for my purposes. For those of you unfamiliar with the story, a poor shoemaker only has enough leather left for one pair of shoes. He cuts the leather and lays it out for stitching the next morning. He awakes, amazed to find the shoes made, and meticulously crafted at that. Elves apparently helped out during the night, soon to bring fame and fortune to the shoemaker.

But what if the elves didn’t exist. What if, instead, the shoemaker was actually making the shoes in his sleep? The idea is not so ridiculous. Rumor has it that Coleridge actually wrote Kubla Khan during a dream, and managed to scribble it down before it faded from his consciousness. As any psychiatrist will tell you, we’re closest to our subsconscious when we’re hovering between sleep and wakefulness. It’s about the only time we get a glimpse into those murky depths.

So let’s say our shoemaker actually makes the shoes in some bizarre bout of sleepwalking. He awakes every morning, to find the shoes nearly perfectly finished. All he needs to do is add the laces and a bit of polish. And the shoes are fair more carefully crafted then he could ever accomplish while awake.

The shoemaker really isn’t aware of where the shoes come from. In fact, as time goes on, and as he receives more and more recognition for the quality of his workmanship, he begins to believe that it’s solely due to the little bit of work he does while he’s awake, threading the laces and adding a little polish. He learns to ignore the 95% of the work that’s done while he’s asleep.

Now, imagine someone comes to ask him why his shoes are so exceptionally crafted. Would he admit the truth and say he doesn’t know? No, pride and genuine lack of knowledge would keep him from saying that. He has no idea what he does while he’s asleep. It’s almost as if someone else did the work for him. His conscious brain would kick in and come up with some perfectly rational but completely untrue explanation. Clotaire Rapaille, in his book The Culture Code, cites an example of this:

In a classic study, the nineteenth-century scientist Jean-Martin Charcot hypnotized a female patient, handed her an umbrella, and asked her to open it. After this, he slowly brought the woman out of her hypnotic state. When she came to, she was surprised by the object she held in her hand. Charcot then asked her why she was carrying an open umbrella indoors. The woman was utterly confused by the question. She of course had no idea of what she had been through and no memories of Charcot’s instructions. Baffled, she looked at the ceiling. Then she looked back at Charcot and said, “It was raining.”

This is what happens in almost every instance of market research. Our buying decisions are like the shoemaker’s shoes. They’re usually quite good, but we have little idea how they came into being.

For most of the history of marketing, we’ve been restrained by the limitations of market research. It’s only recently, through advancements in cognitive psychology and brain scanning technologies that we’re beginning to get a glimpse of what might actually be happening. My next post (tomorrow) why it’s important that we keep trying.

Edison Also Asked: “When Will People Get It?”

First published November 15, 2007 in Mediapost’s Search Insider

Over the past few weeks, my general theme has been “why don’t more people get it?” Why don’t agencies get search. Why don’t CEOs get search? Why don’t more search portals get that it’s the user that determines your success? Why don’t more people get that the world is changing, quickly? What’s with us, anyway?

Well, this week, I gained a little insight; thanks to a paper by Paul David called “The Dynamo and the Computer.” Maybe we just need some time. It’s not the first time this happened. Let me tell you the story of the light bulb.

Lighting Up the Industrial Age

Edison introduced the first practical incandescent light bulb in 1879. The first generating stations in New York and London started their dynamos spinning in 1881. Profound changes were to follow. Productivity was to grow by leaps and bounds.

Factories in the 1800’s were dark, noisy and not particularly pleasant places to spend a day.  At the center sat the steam engine: a huge, hungry and finicky behemoth, connected by an extended system of belts to the operating machinery of the factory. Even the early electrical engines were smaller, cleaner and much more efficient. Electric lighting made 24 hour shifts more practical. The benefits were obvious. Electricity was the ultimate “no-brainer.”

Still in the Dark

But by 1899, almost two decades after the introduction of the light bulb, only 3% of homes were “wired.” And the much-predicted impact on the North American industrial engine would have to wait until the 1920s to take hold. It took a half century for electricity to make much of a difference in America.

You see, technology tends to move fast, but people move slowly. It’s because transition tends to be dependent on many factors. It’s not like the flicking of (quite literally, in this case) a light switch. It’s more like waiting for a long series of dominos to fall into place, each drop contingent on the previous one.

In the case of electricity, significant money had been invested in steam power. You don’t just rip it all out and start over again, no matter how compelling the advantages might be. So factory owners waited for things to break down, and then retrofitted with new electric engines. But even this retrofitting had to wait for the supply of electrical engineers to catch up. In 1899, not many people knew how to design an electrical delivery system. The skill gap had to be eliminated. And this lack of expertise also showed up in less direct ways. America also had to wait for a new generation of factory architects to appear, who could design factories built to be powered by electricity. For every obvious benefit of electrification, there was a long series of factors that had to fall into place first. That’s why it took five decades to turn on the light.

History Repeating Itself

This technology adoption curve has been repeated over and over. The replacement of horsepower with steam power. And more recently, the information technology revolution. We can get as frustrated as we want with the snail’s pace reluctance of many to grasp the realities of the new world, but the fact is, we’re just being human.

Technology adoption usually follows a predictable path: introduction of technology, commercialization of technology, layering the technology onto what preceded it, and finally, throwing out the old completely and building from the ground up to embrace the technology. Each step depends on the step before it. And in every case, legacy investment slows the speed at which we move from one to the other.

If we look at the adoption of Internet technology and compare it to previous technology adoption curves, we’re just starting up the beginning of the long and steep part of the “S” curve. There’s no doubt we’ll get there, but it will take time.