A friend recently asked me if I would talk to the class he teaches at NYU about the trends I thought would affect digital marketing the most over the next ten years. After I composed my lecture notes, I realized that the list might be useful to professionals who are already engaged with digital marketing. So, we’re taking a break from this week’s normal digital trends email to address… Six Things Every Digital Marketer Should Know 1. eSports Interactive games have never been for socially awkward, morbidly obese people living in their parent’s basements. And yet, this inaccurate impression of gamers continues to keep marketers from realizing the awesome potential of eSports and interactive gaming as a platform. It is my sincere hope that Mary Meeker’s latest Internet Trends presentation will finally awaken marketers to the one medium that has the potential to replace the engagement and mass audience of television. Meeker entitles the section on interactive gaming “Motherlode of Tech Product Innovation/Evolution + Modern Learning.” The estimated number or regular gamers worldwide in 2017 is 2.6 billion people. Because of the evolution of interactive gaming technology, these individuals are engaged in global collaborative play, usually with a heavy social media/messaging component. While North America and Europe constitute a large amount of the growth in gamers, the largest growth is happening in the Asia Pac region. While 18-50 years old make up the bulk of the gaming world, the number of people over 50 who play interactive games regularly is equal to the number of under 18 year olds. Meaning, that this is a truly cross-generational medium of a type we haven’t seen since the monopoly of the big three television networks. While gamers are more likely to be men in general, in certain game genres, women outnumber men. Finally, the viewership for the large eSports leagues is growing by almost 20% a year and compares favorably with many professional sports events. Why does this matter? Marketers are always looking for “the next Michael Jordan” – a mediagenic personality with cross generational appeal. It seems likely that the next Michael Jordan might be playing eSports right now. As media fragments, marketers chase the elusive “mass audience” by cobbling together smaller demographics across multiple entertainment channels. But the mass audience exists and it is playing games. Further, interactive gaming is a more engaged form of media consumption than the old TV model. It is ridiculous that marketing dollars continue to over index for television and magazines and interactive gaming isn’t even measured as a media category. I’ve been saying this for years: This is the marketing medium for the 21st Century. 2. User Experience Design User experience design is a new commercial art form and, as marketers, we’re missing the boat. Right now, the practice of user experience has been reduced to simplistic and reductionistic techniques like sitemaps and wireframes. At best, some companies are using heat map technology to tell them what areas users gravitate towards on their interfaces. But the design itself of most digital interfaces has been reduced to two dimensional print design with a few clickable buttons. What’s wrong with this approach? Well, a graphical user interface or GUI is just a convenient fiction. It doesn’t really exist. It is just a representation of an artificial environment that makes users comfortable dealing with complicated, digital technology. Your computer doesn’t really contain a trash can that gets emptied. It’s just an easier way for most people to conceive of deleting a file than typing it into the command line. Websites also don’t really exist as physical environments. There are no pages, no funnels, no buttons. A website or an app is just a graphical representation of a bit of functionality or content. None of it actually exists. Why does this matter? If it doesn’t really exist, if it’s only a story we’re telling each other, then there is no reason to tell such boring stories. Yes, users benefit from consistency of design across sites and apps, but user experience design has taken that as an excuse to create a sea of sameness. Remember, that we are now able to create experiences in four dimensions (including time) and that those experiences can be fully interactive to our users. This is, quite simply, an unprecedented opportunity for graphic experimentation and we’ve reduced it to responsive boxes on a grid. If you are creating a website or an app, boil your challenge down to the bare bones of the content or functionality you are offering your user. Then build up from that foundation and don’t let convention and habit blind you to new possibilities. Tell better stories. 3. Backpropagation Machine learning is a highly useful technique for machine-based pattern recognition that is housed within an unholy tower of bullshit. First, many people insist on calling it Artificial Intelligence or AI. It is not intelligence. Even machine learning is a misleading phrase since learning – as we understand it in humans and animals – is not taking place. Finally, the whole thing is supposedly based on something called a neural network, only the nodes of that network do not even vaguely resemble actual neurons. So we have a powerful and useful technology that is being actively misrepresented by marketers in a haze of puffery and grandiose language. Far be it for me to criticize other digital marketing professionals for engaging in a little casual bullshit in the pursuit of billings. However, there is an actual technology underneath all of this hot air and, like most technologies, it is good at some things and bad at others. The thing it is best at is pattern recognition. Why is this important? Many people believe that pattern recognition is one of the fundamental pillars of human intelligence. When we use language, we are recognizing and manipulating patterns of noise or symbols (letters or characters). When we see an object and recognize it as a dog or cat, we are using pattern recognition. Narratives are also patterns and so all of marketing is based on patterns. One can imagine that a technology capable of recognizing and manipulating patterns might have substantial implications for our industry. Unfortunately, people are starting to treat machine learning like a “secret sauce” to be sprinkled over everything in order to make it more “digital” and “objective.” But machine learning is only as good as the data that the system is trained on. If that data is biased, then the process will reflect that bias. If that data is mislabeled, then the whole thing is pretty much worthless. Marketing is always slow to recognize the snake oil salesmen in our midst. However, clients do eventually wonder why their huge technology investment isn’t yielding a return on investment. In the long term, it’s worth understanding what you’re selling and only sell it if you think it could be effective. If you really want to understand machine learning, you need to understand backpropagation. Let’s say you’re trying to train a neural network to recognize the difference between cats and dogs. (I don’t know why, maybe your target audience is people with dogs.) You have a large dataset of images of people with dogs and people with cats. You have labelled that data so that after the neural network has processed the images you can compare the neural network’s guesses with the anticipated result. So if the network guesses that it’s a person with a cat, but it’s really a person with a small dog, that returns as an error. Because you are a smart person, your initial neural network has random weighting. Meaning that you aren’t introducing bias into the system to begin with. You’re letting the system optimize itself. Once the system yields an error (that’s not a cat, it’s a small dog) the error propagates backwards through the system and readjusts the weighting. The more errors the system automatically corrects through this feedback mechanism, the better the eventual system will be at recognizing cats and dogs. That is backpropagation and, when you understand that, you understand more about where machine learning will be most effective. 4. Augmented Reality Like every science fiction-obsessed nerd who dreams of using the interfaces in Minority Report, I was devastated to learn about the Magic Leap demos. Magic Leap (for you non-obsessives) is a secretive augmented reality company that has attracted heavy investment from a who’s who of Silicon Valley venture capitalists and companies. As the company raked in billions of dollars in seed money, they refused to reveal much about what they were building. The founders even located their offices in Florida, far from established technology centers to keep their technology from leaking to the tech press. Only a few demos were allowed to leak out and those demos sent technologists (like yours truly) salivating with excitement. And then the news came out that the demos had actually been created by New Zealand based special effects house Weta Digital and did not reflect the current state of the technology. While I have no doubt that Magic Leap and Google and Apple and Microsoft are actually at work on amazing AR technology, we are clearly far away from a large scale commercially viable product. The goggles are too heavy and the need to map AR onto the real world has proven to be a substantial stumbling block. If I want there to be a goblin sitting on your desk, AR technology needs to recognize that your desk is an object and that it has a horizontal surface capable of holding a goblin. Then, that goblin needs to be realistically imposed on the scene in exactly the right way so it appears to be sitting and not floating. This requires a massive amount of computing power and is not currently viable in a Google Glass-like interface. Why does this matter? If you ignore technology because it is too expensive and requires too much computing power to be commercially viable, you would have missed the last 70 years of technological innovation (including television.) Eventually, AR and VR (which are really just different ways of looking at the same underlying technology) will be commercially viable. Smart marketers will begin to experiment with the early iterations of this potentially world-changing technology so that when this new media delivery mechanism is commercially viable, they are not left behind. And it is a mistake to look too hard at science fiction as a guide to what these technologies will eventually look like. Don’t picture the holodeck. Picture your office without any computers on the desks. Also, and I hate to say this, the earliest commercial applications are likely to be pornographic. Make of that what you will. 5. Driverless Vehicles I know, I know, this doesn’t sound much like marketing technology. Here’s why I am including it in this list: marketers have always shaped their efforts around the society in which they live and work. The old Burma Shave ads which stretched out the marketing messages over several billboards on a highway are one example of how our car-centric society has shaped marketing. When people imagine driverless vehicles, they tend to picture something like an uber car with no driver. But that isn’t even close to what we will see eventually. First, most vehicles are oriented around the comfort and convenience of the driver, since they are likely to be making the purchase decision. Second, most vehicles are individually owned. But that feels like an odd decision if all vehicles are driverless. Most people didn’t feel the need to buy their own railroad engine 120 year ago. Third, the cityscape will change radically once driverless vehicles are on the road. With an accident rate approaching zero and the subsequent fall in insurance costs, driverless cars will become the majority of cars on the road very quickly. Once vehicles are driverless, there is no need for stoplights or traffic signs, parking places or traffic lanes. Cities will transform in ways we have not seen since the advent of the automobile. Why does this matter? Driverless vehicles mean more free time. Commuters will find themselves free to work, enjoy some form of media or nap. Commercial radio, which has limped along on the residue of commuter traffic will dry up. People will look for choices. Without the need to focus on the road, the highway billboard is an absurd contrivance. It will exist primarily for people who experience motion sickness and need to focus on the road ahead. And many of today’s biggest marketers (Ford, GM, Chrysler, Geico) will have different targets and different messages in the future. That is, assuming they exist at all. 6. The real cost of “free” Years ago I had a meeting with a prospective client. She was starting a new company with a product aimed at baby boomers who were caring for elderly parents. This was a big potential market at the time. She asked how I would market to this audience and I started talking about where I thought she should advertise and what messages I thought she should communicate. “Ha!” she exulted. “I was warned you would try to get me to spend money on advertising. But I’ve asked people I trust and they tell me that all marketing is moving to social media and social media is free!” When I gently informed her that social media marketing also costs money, she angrily insisted that I was trying to gouge her. Why she imagined I would or should do work at her preferred price (free) I cannot guess. I was not surprised when her company failed. Supposedly “free” marketing always has a hefty price. Adidas recently had to learn a similar lesson. I should say, that I like Adidas as a company. Adidas entered into a sponsorship of the Boston Marathon which certainly makes a great deal of sense for a company that makes running shoes. I’ve worked on this type of sponsorship deal before. You typically get some signage at the start and finish of the marathon. Perhaps additional signage along the course at prominent locations like water stations. You get prominent placement in the program and on the event website. Your logo goes on the poster. And you typically get a branded tent in the athlete area for sampling and brand goodwill. Typically, far, far down on the list of things you get as part of the sponsorship is the opportunity to email participants before and after the event. It’s considered something of a “gimme”, since it doesn’t cost the organizers anything. It’s also pretty low engagement compared to all the other aspects of the sponsorship. The day after the Boston Marathon, Adidas sent out an email which was titled “Congratulations on Surviving the Boston Marathon.” This was an unintentionally tasteless marketing message given the recent history of the marathon in Boston. Why does this matter? I think marketers make a mistake when they accept free marketing because the cost blinds them to the potential downside of a marketing message. Any marketing message needs to be evaluated so that the potential upside outweighs the potential downside. Imagine for a moment the potential upside of the Adidas email to Boston Marathon participants. Do they sell a dozen pairs of shoes? Now compare that to the downside of delivering an insensitive marketing message to the entire city of Boston. Yes, cost of media is also a downside, one that budget-conscious marketers are quite focused on. But free media may conceal other downsides that marketers are ill-equipped to evaluate. A marketing organization is made up of human beings and human beings tend to value what they pay for. Free add-ons like the post-race email don’t receive nearly the scrutiny of the higher profile impressions. Adidas is not alone in this. For the last couple years we’ve seen experienced marketers make screamingly-bad mistakes on social media on a weekly basis. If I may be so bold, I would propose the following formula to evaluate every marketing message, even (especially) the free ones: E(x | P(x)) - E(y | P(y)) ≥ 0 Where x is the potential benefit and y is the potential disaster.