Archive for the ‘Targeting’ Category
I just spoke at the PIPA Conference on Security on Behavioral Targeting with James Bond. Yes, that’s his name.
It was interesting to speak to so many people in one room whose job is to control, limit, and say “no.” I asked them to all introduce themselves to the person next to them and tell that person one thing about themselves that they would not normally tell someone they just met.
“Why did you do that?” I asked. Why, because I asked them too.
The internet works much the same way. Sites are constantly asking you for information which you divulge; names, email, your location etc… Realize that even just clicking on a link, you are saying something about yourself. You divulge it because each time its not a lot of information, so you do not feel threatened, do not feel violated. The problem with the internet is that we now have the ability to scoop up each little tidbit about you to create a profile of who you are, what you may like, what you are interested in. In fact, we may be able to construct a more complete profile of your interests than you even know. That is because it will be based on what you actually do, as opposed to what you think you are interested in.
I will go so far as to question the very reason for this conference. Are we barking at moon? Can we control privacy in a free society? And what about Behavioral Targeting is SO threatening to it? Yes, a little extreme, without so semblance of guidelines it would be complete anarchy. But it is really only an illusion of order now. But that is essential. We all buy into that illusion, and that’s what creates the order in the chaos.
Never before have we had the technology to stretch beyond our own imagining. Computer programs approaching that of Free Will in their decision making. Algorithms, intricate in design, yet simple compared to the vast sea of data they crawl.
You used to play hide-and-seek as a child. Over time if you played with the same people they would discover your favorite hiding places. You are, much to your chagrin, much more predictable than you would like to believe. And so is your online persona. That is what the new technology enables. Over time, it uncovers all the little places you hide. It’s as if you are playing hide and seek, and every-time you go to one of your hiding places there is a slip of paper. You pick it up, open it, and it says “I’ve found you.” In real life this would be a surreal experience, and one that causes you consternation. But what if, every-time you went to one of your hiding places, instead of a note, there was something useful. A flashlight in that dark hallway so you could light your way? A book you wanted to read in the closet so you could read while you were waiting? It’s the difference between active acknowledgment with the note, and passive rewarding with the useful items. Providing you what you didn’t even know you may need but have the proclivity to respond to. You didn’t ask for it. Most would find that very helpful. That is what Behavioral Targeting enables.
That is the issue we are facing, and this is the decision that we all make. That is the issue of this conference. How much do we want intrusions into our illusions of privacy for convenience? Do the benefits outweigh the risks?
This is not new. Grocery stores have been assembling profiles on shoppers for years with “discount cards.” They do not discount merchandise because they are altruistic. They are using it to make more money. How many of you have a Safeway card? That discount card… which tracks every purchase you have ever made… ever. Now, how many of you buy alcohol, cigarettes, or condoms there? High fat foods? Hmmm… what if that data was used by an insurance company? But the data of these organizations is used vertically, to improve their store sales, to put products where people will purchase them, to create need states for particular products. Credit Card companies know more about your lifestyle and purchases than you are probably comfortable with, yet we use them for convenience.
Catalog companies are one of the biggest resellers of your name, address, and profiles in the world. If they could not do it they would not be able to survive economically, and yet, convenience. In the end people, like with the hide & seek example are habituators. Habitual pattern developers, and followers. That is what enables marketers to be so successful.
1.0 technology was not intelligent about the data, it was not really Behavioral Targeting, or it was, but the data sets were limited. People became very vertically sliced representations of themselves to that individual company. But is a more holistic view of the individual increasing the accuracy for individual companies? No. Its value is to that of the advertiser.
Ad Networks in a vast sea of non-specificity where improvement in ad targeting allows for ongoing monitoring of consumers whose behaviors shift over time. Once you’ve bought the house, your need for the mortgage decreases, however, you shift into a different category of home improvement and furniture selection, insurance policies, and school districts.
There is no security online… accept it, and move on.
Do we want the Internet to be more intuitive? or less so?
It’s a global issue trying to be controlled locally by governments. How do you know if that site that appears Canadian is even located in Canada? Anyone can register a .ca domain, regardless of what some think. I’ve done it. The domain name system is not an indicator of your security, or the laws that bind them.
But the attitudinal shift difference between generations is faster than that of employers shifting to adapt. GenY and Millennials do not have the same illusions of security, nor do they really have that much credit to destroy yet. However, what they do not seem to do is find delineation’s in their private and public selves. This may come back to haunt them, but they are not past, or future. More so than other generations they seek to be conscious in the present.
There are a number of people reading this who visit porn sites. I know, shocking. Most of you are smart enough to use aliases and email addresses you use specifically for this purpose. But MOST of you do not mask your IP address. If you think US laws on security are bad, do you think that a website in Russia cares one iota about your data or security? In fact, many of these sites exist specifically to mine your information for identity fraud. Rings of sites waiting to connect hat IP address with real information.
When we talk about DPI (Deep Packet Inspection) and other technologies I really think senators want the “opt-in” because of their own duality. What if Governor General was found out to be a frequent visitor of such sites? But opt-in is always the first step… it dissipates over time with the benefits. Remember, this is the same argument that happened over email. Now checkboxes are routinely checked YES as opposed to NO. You opt-ed in, but are often unaware that you did.
What about real sites? Lawyers construct complicated legal agreements that no one reads online, specifically crafted to allow for change and variability. All they do is give the illusion of control to data. Most are written so that the legal agreements do not survive the company if sold. Self serving documents designed to protect the company, NOT the individual.
What’s the harm?
YOU are actually in control of the data YOU post, the way YOU surf, what YOU do. I will argue tat as soon as you go out of the confines of your own home, either physically or virtually, you lose ceratin rights to contain information about yourself. If you click on the porn link, that was in the public domain outside of your home. If the consumer does not take care with their data, why should companies respect it? That is a serious question. Why? Because the individual is often ignorant of the dangers of what they post online UNTIL they are affected by it.
What role does government have to play? Actually a very limited one in my opinion. No one government can protect your data. All they can do is set laws to provide the illusion of your security. Provide you with a recourse of action, but usually the damage has already been done. If public, it often does not matter what the truth is, it matters what the perception of that truth is.
Look at airport security for an illusory example. You all realize it’s an illusion right? You get to the point of being cavity searched just to fly a plane, yet they load on tons of cargo. 1% is checked if we’re lucky, a check mark on a manifest by an overworked employee unhappy with their job. The loopholes in everything is people. Society is not a clean organized mechanical being – well unless you’re in Singapore. It’s messy, chaotic and subject to the moods and whims of its populace.
In the global age, we need a global governing body for information, and no one should have that kind of power. With 54 of the top 100 economies in the world being corporations, government is increasingly limited in it’s purview to control information.
Unless you are the government actually looking to control the populace, as with The US Patriot Act (I love that name… as if I am not a patriot if I don’t support it) which allows the request of any information it wants in the name of “terrorism” without the individual being notified.
Should we be moving to an arena of less governmental laws and control, or more?
What are the real risks? Identity theft, fraud. There is a profound aversion to you not being you. That someone could take it, assume your identity. It is a loss of self, the most profound societal violation, but only if people let it be. You are not a credit card number, not your name or your credit report, your registration data.
What do you expect to be private? and can you let go of that resistance? Embrace change? The old world we once knew is gone. To remain anonymous you must now truly be off the grid.
Do we have the time to adjust to this new world order?
Me, my contact information… Anyone could find all of this online with a little effort, but, at any one destination it would only be one or two bits of it. This is what Behavioral Targeting enables. Now, if they use this data to deliver better content to me that’s wonderful. I get to live my life more efficiently… more likely I’ll just get more ads for dating sites, a new mortgage, or a car I don’t need.
Sean X Cummings / 40 / Male / Single / Liberal / Stroke survivor / Buddhist / OM / Marketing / Advertising / Writer / Facebook / No MySpace / Childhood Europe / College Cornell / Dual Citizen USA / Ireland / High-energy particle physics / drives Ducati / Skydiver / Ask.com / American Express / C-K / Leap / Startups / Tattoo’s / Wears toe polish / Sensuality practice / Drinks alcohol / Ex-Hacker / Burning Man / Visits porn sites / Reads SciFi /BSG / Dexter / House / Weeds / Heroes / Independant Film… oh yeah… and I hate penguins
The long tail has been the mantra of online marketing, but is it really the right paradigm? Find out why the prevailing wisdom may be flawed.
Spend a minute with anyone working in interactive and you’re bound to hear at least a vague reference to the long tail — a theory put forward by Chris Anderson in his book by the same name. But for all the talk, interactive is still grappling with two fundamental questions: What does the long tail really mean, and is it something that you should be pursuing?
The long tail has different meanings whether you are in manufacturing or advertising. But it basically refers to the idea that a small group of product offerings or keyword lists covers the majority of the market opportunity. It’s really the 80/20 rule. If you only have to make a certain number of products to cover 80 percent of the market, it would require a lot more effort to go after that remaining 20 percent. Often that 20 percent is too much to justify the costs, begging the question: is it worth it?
The same holds true for keywords. It’s likely that a small number of keywords bring in 80 percent of your traffic or sales. But what about the remaining 20 percent?
Before automation on both sides, that portion of the market was impossible to go after efficiently. Now that automation makes it possible, the question remains: should you?
This article will speak to the product side. Later, I will tackle the other side of the equation — SEM.
The long tail reality
The internet was going to enable thousands, no, tens of thousands of niche offerings. And those offerings would be so highly-customized to speak to consumers’ need states that you’d be optimizing and reaching the maximum market opportunity. The reality is something different. There are arguments on both sides as to the value, but it is often a perceptive conclusion. It all depends on what data set you’re looking at, and whether you are the big brand looking to expand product offerings or the small brand concentrating on a niche.
There was an intriguing article in the Harvard Business Review about whether you should invest in the long tail. It’s a great article, but it draws some strange conclusions as to next steps. The article speaks to what is essentially the 80/20 rule, using various published studies. Theory is great and all, and citing other studies is often quite impressive to bolster your point, as this article did. However, the devil is in the details.
Example beyond the data
First, the article draws several of its conclusions based solely on the sales data. Tsk, tsk. “But Sean X, data doesn’t lie!” No, it tends not to, but it does not tell the whole story, or even remotely deal with the dynamics of intent that drive those purchases — the purchase cycle. It is like looking at a myopic funnel of the process and extrapolating conclusions. And that is a dangerous road to walk.
What that article misses, is the “phenomenom of choice.” I will provide several examples:
A well known 24-hour grocery store found that even though it incurred 20 percent of its expenses from 10 p.m. to 6 a.m., it only made 6 percent of its sales during that period. It decided to change its hours from 24 hours per day to closed from 10 p.m. to 6 a.m. Their logic was that they would save 20 percent of overhead for only a 6 percent profit cut. But sales actually dropped 30 percent for a net loss of 10 percent. Why? The phenomenon of choice. By changing from the ultimate convenience, 24/7, the perception in the consumers’ minds was that the store was less convenient. “Were they open ’till 9 or 10? I don’t know. I’m going to the other place.”
Just looking at sales data captures the end of the process, the result. It does not deal with the dynamics of what generated it. Logic based on data, often has illogical outcomes.
Campbell’s Soup often has half an aisle in the supermarket dedicated to its red and white cans. However, Cream of Mushroom and Tomato account for a significant slice of the total profits of that aisle. Should Campbell’s just make fewer soups, use less shelf space, and get rid of the dogs? Of course not. They understand that when the consumer sees a swath of red and white cans, Campbell’s is soup. They peruse and then choose their Cream of Mushroom, Broccoli Cheese and Tomato soups. The phenomenon of choice.
The long tail often wags the short tail. Be careful to study what the impact will be with consumers on the perception of your brand, and not just sales data.
Herman Miller advertised its Resolve furniture line, even though it accounted for a small fraction of its sales at the time. Consumers drawn in by the design aesthetic of that line of products often made more practical choices when it came right down to it. But it was that line that brought them in the door. Without it they were going to be out of the buyers’ consideration set. And that is the difference. If you’re just looking at the “sales” data, you’re not measuring the dynamics of consumer intent. And that is the major flaw in the HBR article criticizing the long tail.
What gets people in the door is often not what they eventually buy. Product differentiation is often key in consumer choice. Sexy products sell the less attractive and less expensive ones.
Even though 90 percent of the movies you rent at Netflix are blockbusters, the phenomenon of choice is why you are a member. If they only carried the top 100 titles would you join? No, because you want the choice, even if you’re not going to use it. If they scaled back just to the blockbusters, the consumers would abandon the service. What the article does correctly point out is that resource allocation of the long tail is key.
The phenomenon of choice drives membership. Be careful when eliminating choice if you are a membership based service.
So what should you do if you are a manufacturer or brand?
When considering the long tail, manufacturers and brands should ask the following questions:
1. What are the resources required to make niche products or carry them in your catalog?
2. Will those niche products bring consumers in the door, even though they’ll buy more conventional offerings from you?
3. Do those long tail products provide a marketable point-of-differentiation in your competitive set that can be leveraged in advertising?
4. What’s the loss in not making them? Not the hard numbers, but equating consumer shift in mindset because of their elimination.
The “hard” numbers in sales data sets are often a driving force behind many decisions in business. They are a known quantity. Myopic managers often use hard data as their decision making tool because even if the result is negative they can point to hard justifications for the decisions. It is the result of corporate structures that do not reward success but do punish failure — a review structure where as long as no one has anything negative to say the person gets promoted. Elevation through mediocrity. That mentality is the hallmark of marketers coming from a consulting background who do not understand the fuzzier “gut” decisions made by seasoned marketing professionals.
However, that gut decision is not really a gut decision at all, it is usually based on consumer research, psychographic studies, survey data, focus groups, observation and a deeper understanding of the dynamics of consumer intent. The problem is that most of the time, decisions based on that data do not have the concrete numbers to fall back on. It’s fuzzy. That so-called marketers’ instinct, the ability to consistently call it right on those decisions, comes from something else — knowledge and experience. Marketers who operate in that manner often get the comments “How did you know?” when a program is a success. It can’t be taught in schools; it has to be learned in the real world. But what makes them that good is often intangible.
What can you do if you are stuck in such an organization? Well, if your company is run by CFO types, not much. You’re kind of, what’s the word I’m looking for? Oh yeah, “screwed.” Hard data is their friend. A way to approach that mentality is with as much data as you can muster. For example, you might run Netpromoter score data if you have an online presence. That data, although based on consumer opinions, can often be tied directly to sales fluctuations. Also, econometric modeling has come a long way. There are several companies who can help provide all the inputs that affect the sales data, instead of just myopically concentrating on the sales output.
Marketers who operate with instinct, and a more thorough understanding of the dynamics of consumer intent, understand that it is the future trends that are important. Those who base decisions only on concrete data are only looking at the past, hoping for a repeat performance. And the market is often not dynamic.
In the end, don’t believe what the experts who write the books say, don’t believe what a competitor’s data says, don’t even believe anything I write here. Your company is unique. It has challenges that are based on its people, its manufacturing process and its corporate structure. But what if a competitor in your space is getting better traction, making more money and growing even though you think their process is flawed — they’re just lucky, right? No, they’re thinking the same thing about you. Only difference is, they’re right.
ranty rant signing off…