157 Shades of Grey…

Design is important. Thinking through how people will respond to the aesthetics of your product is an admirable thing. I remember once having the pleasure of sharing a stage with JetBlue’s VP of Marketing – Amy Curtis-McIntyre. She was explaining how important good design was to the airline’s overall marketing strategy. A tremendous amount of thought went into the aesthetics of all their printed materials – even those cards explaining the safety features of the airplane that none of us ever read. But on JetBlue, not only did passengers read them – they stole them because they were so cleverly designed. Was this a problem for management? Not according to Amy:

“You know you’re doing something right when people steal your marketing shit”

So, I’m a fan of good design. But according to a recent story on Fastcodesign.com, Google is going at least 156 shades too far. They seem obsessed with color – or – at least, testing for colors. The design team for Google’s new home assistant – the Mini – had to pick three different colors for the home appliance. They wanted one to make a personal statement and apparently that statement is best made by the color “Coral.” Then they needed a color that would sit unobtrusively next to your TV set and that turned out to be “Charcoal.” Finally, they needed a “floater” color that could go anywhere in the house, including the kitchen. And that’s when the design team at Google may have gone off the tracks. They tested 157 shades of grey – yes – 157 – before they settled on “Chalk,” which is said to be the most inoffensive shade imaginable. They even worked with a textile firm to create their own custom cloth for the grill on top.

That beats Google’s previous obsessive-compulsive testing disorder record, set by then VP of Search Marissa Mayer when she ordered the design team to test 42 different shades of blue for search links to see which got the most clicks. At Google, good design seems to equal endless testing. But is there anything wrong with that?

Well, for one thing, you can test yourself into a rabbit hole, running endless tests and drowning in reams of data looking for the optimal solution – completely missing global maxima while myopically focused on the local. Google tests everything – and I mean everything – truly, madly and deeply. Even Google insiders admit this penchant for testing often gets them focused on the trees rather than the forest. This is particularly true for design. Google has a long history of obsessively turning out ho-hum designs.

Personally, when it comes to pure design magic, I much prefer the Apple approach. Led by Steve Job and Jon Ive’s unerring sense for the aesthetic – it’s hard to think of a longer run of spectacular product designs. Yes, they too sweated the small stuff. But those details were always in service of a higher vision – an empathetic, elegantly simple, friendly approach to product design that somehow magically connected with the user, leaving that user somewhat awed and consistently impressed. One might quibble with the technology that lies inside the package, but no one has put together a more beautiful package that the Apple design team at the height of their powers.

When you look at a Google product, you have the result of endless testing and data crunching. When you look at a classic Apple design, you sense that this came from more than simple testing. This came from intuition and creativity.

 

Trust No One, Trust Nothing

In just one day last week looking at the headlines on MediaPost – two different articles mentioned a lack of trust – a lack of trust in contextual ad placement and a lack of trust in audience measurement data. But our industries trust issues go far deeper than just those two instances. Article after article cite an erosion of trust and the spreading of relational fault lines in every aspect of the industry.

The question of the day is “Where did the trust go? The follow up question then becomes, “What do we mean by trust?”

That is a difficult question. Trust is a word with many, many meanings. Over 20 years ago, University of Minnesota business professors D. Harrison McKnight and Norman L. Chervany wrote an extensive review to answer just that question. In it, across the many constructs of trust, they identified four dimensions: benevolence, integrity, competence and predictability. But not all these dimensions are required in all applications of trust.

First of all, there are two broad categories of trust: structural trust – trust in a system – and interpersonal trust – trust in a person. In their analysis, McKnight and Chervany looked at six constructs of trust that can apply in different situations. For the sake of this discussion, let’s focus on two of these:

“System trust: the extent to which one believes that proper impersonal structures are in place to enable one to anticipate a successful future endeavor.”

And…

“Situational Trust – meaning that one has decided to trust without regard to the specific persons involved, because the benefits of trusting in this situation outweigh the possible negative outcomes of trusting.”

What trust that did exist in marketing what an outcome of these two constructs. Both tend to apply to the structure of marketing, not the people in marketing. The headlines I cited earlier both pointed to a breakdown of trust on the system level, not the personal level. Now, let’s look at those four dimensions as they apply to structural trust in marketing. No one has ever accused marketers of being overly benevolent, so let’s set that one aside. Also, I would argue – strenuously – that marketers today – including those at agencies – are more competent than ever before. They have been mostly successful at turning marketing from an arcane guessing game that paraded as art to an empirically backed science. So a lack of competency can’t be blamed for this trust breakdown. That leaves integrity and predictability. I suspect there’s a compound relationship between these two things.

The reason we’re losing structural trust is that marketing is no longer predictable. And this lack of predictability is triggering a suspicion that there has been a corresponding lack of integrity. But the unpredictability of marketing is no one’s fault.

Marketing today is analogous to physics at the turn of the last century. For 200 years the universe had been neatly ruled by Newton’s Laws. Then physicists started discovering things that couldn’t be so neatly explained and the Universe became a place of Uncertainty Principles, Schrödinger’s Cat and Strange Attractors. Everything we thought was predictable in all situations suddenly become part of a much bigger – and more complex – mystery.

Similarly, mass marketing could run by Newton-like laws because we were dealing with mass and weren’t looking too closely. Apply enough force to enough people with enough frequency and you could move the needle in what seemed like a predictable fashion. But today marketing is a vastly different beast. We market one-to-one and those “ones” are all inter-connected, which creates all types of feedback loops and network effects. This creates complexity – so predictability is as dead at the afore-mentioned Schrödinger’s Cat (or is it?)

I don’t think this comes as news to anyone reading this column. We all know we’re being disrupted. I think we’re all beginning to understand the challenges of complexity. So why don’t we just accept it as the new normal and continue to work together? Why are clients feeling personally betrayed by their agencies, market research firms and ad delivery platforms? It’s because our brains aren’t very nuanced when it comes to identifying trust and betrayal. Brains operate by the “when you’re a hammer – everything looks like a nail” principle.

Rationally, we understand the different between interpersonal trust and situational trust, but we have to remember that our rationality is reinforced by emotional rewards and cautions. When we’re in a trusting relationship – or system – our ventrial striatum, medial prefrontal cortex and caudate nucleus all perk happily along, priming our brains with oxytocin and pushing all the right reward buttons. But whether it’s a person or a situation that betrays our trust, the same neural mechanisms fire – the insula and amygdala – creating feelings of frustration, fear, anger and resentment.

Now, none of this is the fault of anyone in marketing. But humans work on cause and effect. If our marketing is not working, it’s easier to assign a human cause. And it’s much easier to feel betrayed by a human than by a system.

 

I, Robot….

Note: No Artificial Intelligence was involved in the creation of this column.

In the year 1942, science fiction writer Isaac Asimov introduced the 3 Rules of Robotics in his collection of short stories, I, Robot..

  1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
  2. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
  3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws.

Asimov had the rules as coming from the Handbook of Robotics, 56th Edition, 2058 A.D. What was once an unimaginably distant time in the future is now knocking with increasing intensity on the door of the present. And Elon Musk, for one, is worried. “AI is a fundamental risk to the existence of human civilization.” Musk believes, Rules of Robotics or no, we won’t be able to control this genie once it gets out of its bottle.

Right now, the genie looks pretty benign. In the past year, the Washington Post has used robot reporters to write over 850 stories. The Post believes this is a win/win with their human reporters, because the robot, named Heliograf, can:

  • Cover stories that wouldn’t have been covered due to lack of human resources
  • Do the factual heavy lifting for human reporters
  • Alert humans to possible news stories in big data sets

So, should we fear or cheer robots? I think the Post’s experiment highlights two areas that AI excels at, and indicates how we might play nice with machines.

For AI to work effectively, the dots have to be pretty well sketched out. When they are, AI can be tireless in scouting out relevant facts and data where humans would tend to get bored easily. But humans are still much better at connecting those dots, especially when no obvious connection is apparent. We do it through something called intuition. It’s at least one area where we can still blow machines away.

Machines are also good at detecting patterns in overwhelming amounts of data. Humans tend to overfit…make the data fit our narratives. We’ll come back to this point in a minute, but for now, let’s go back to intuition. It’s still the trump card we humans hold. In 2008, Wired editor Chris Anderson prematurely (and, many believe, incorrectly) declared the Scientific Method dead, thanks to the massive data sets we now have available:

“We can analyze the data without hypotheses about what it might show. We can throw the numbers into the biggest computing clusters the world has ever seen and let statistical algorithms find patterns where science cannot.”

Anderson gets it partly right, but he also unfairly gives intuition short shrift. This is not a zero sum game. Intuition and A.I. can and should play nicely together. As I mentioned a few weeks ago, human intuition was found to boost the effectiveness of an optimization algorithm by 25%.

Evolutionary biologist Richard Dawkins recently came to the defense of intuition in Science, saying:

“Science proceeds by intuitive leaps of the imagination – building an idea of what might be true, and then testing it”

The very human problem comes when we let our imaginations run away from the facts, bending science to fit our hypotheses:

“It is important that scientists should not be so wedded to that intuition that they omit the very important testing stage.”

There is a kind of reciprocation here – an oscillation between phases. Humans are great at some stages – the ones that require intuition and imagination -and machines are better at others – where a cold and dispassionate analysis of the facts is required. Like most things in nature that pulse with a natural rhythm, the whole gains from the opposing forces at work here. It is a symphony with a beat and a counterbeat.

That’s why, for the immediate future anyway, machines should bend not to our will, but to our imagination.

The Assisted Reality of the New Marketer

Last week, MediaPost’s Laurie Sullivan warned us that the future of analytical number crunchers is not particularly rosy in the world of marketing. With cognitive technologies like IBM’s Watson coming on strong in more and more places, analytic skills are not that hot a commodity any more. Ironically, when it comes to marketing, the majority of companies have not planned to incorporate cognitive technologies in the near future. According to a report from IBM and Oxford Economics, only 24% of the organizations have a plan to incorporate CT in their own operations.

Another study, from Forrester, explored AI Marketing Readiness in Retail and eCommerce sectors. The state of readiness is a little better. In these typically forward thinking sectors, 72% are implementing AI marketing tech in the next year, but only 45% of those companies would consider themselves as excelling in at least 2 out of 3 dimensions of readiness.

If those numbers seem contradictory, we should understand what the difference between cognitive technology and artificial intelligence is. You’ll notice that IBM refers to Watson as “cognitive computing.” As Rob High, IBM’s CTO for Watson put it, “What it’s really about is involvement of a human in the loop,” and he described Watson as “augmented intelligence” rather than artificial intelligence.

That “human in the loop” is a critical difference between the two technologies. Whether we like it or not, machines are inevitable in the world of marketing, so we’d better start thinking about how to play nice with them.

 

I remember first seeing a video from the IBM Amplify summit at a MediaPost event last year. Although the presentation was a little stilted, the promise was intriguing. It showed a marketer musing about a potential campaign and throwing “what ifs” at Watson, who quickly responded with the almost instantly analyzed quantified answers. The premise of the video was to show how smart Watson was. But here’s a “what if” to consider. What if the real key to this was the hypotheticals that the human seemed to be pulling out of the blue? That doesn’t seem that impressive to us – certainly not as impressive as Watson’s corralling and crunching of relevant numbers in the blink of an eye. Musing is what we do. But this is just one example of something called Moravec’s Paradox.

Moravec’s Paradox, as stated by AI pioneer Marvin Minsky, is this: “In general, we’re least aware of what our minds do best. We’re more aware of simple processes that don’t work well than of complex ones that work flawlessly.” In other words, what we find difficult are the tasks that machines are well suited for, and the things we’re not even aware of are the things machines find notoriously hard to do. Things like intuition. And empathy. If we’re looking at the future of the human marketer, we’re probably looking at those two things.

In his book, Humans are Underrated, Geoff Colvin writes,

“Rather than ask what computers can’t do, it’s much more useful to ask what people are compelled to do—those things that a million years of evolution cause us to value and seek from other humans, maybe for a good reason, maybe for no reason, but it’s the way we are.”

We should be ensuring that both humans and machines are doing what they do best, essentially erasing Moravec’s Paradox. Humans focus on intuition and empathy and machines do the heavy lifting on the analyzing and number crunching. The optimal balance – at this point anyway – is a little bit of both.

In Descarte’s Error – neurologist Antonio Damasio showed that without human intuition and emotion – together with the corresponding physical cues he called somatic markers – we could rationalize ourselves into a never-ending spiral without ever coming to a conclusion. We need to be human to function effectively.

Researchers at MIT have even tried to include this into an algorithm. In 1954, Herbert Simon introduced a concept called bounded rationality. It may seem like this puts limits on the cognitive power of humans, but as programmers like to say, bounded rationality is a feature, not a bug. The researchers at MIT found that in an optimization challenge, such as finding the optimal routing strategy for an airline, humans have the advantage of being able to impose some intuitive limits on the number of options considered. For example, a human can say, “Planes should visit each city at the most once,” and thereby dramatically limit the number crunching required. When these intuitive strategies were converted to machine language and introduced into automated algorithms, those algorithms got 10 to 15% smarter.

When it comes right down to it, the essence of marketing is simply a conversation between two people. All the rest: the targeting, the automation, the segmentation, the media strategy – this is all just to add “mass” to marketing. And that’s all the stuff that machines are great at. For us humans, our future seems to rely on our past – and on our ability to connect with other humans.

Is Google Slipping, Or Is It Just Our Imagination?

Recently, I’ve noticed a few articles speculating about whether Google might be slipping:

Last month, the American Customer Satisfaction Index notified us that our confidence in search is on the decline. Google’s score dropped 2% to 82. The culprit was the amount of advertising found on the search results page. To be fair, both Google and search in general have had lower scores. Back in 2015, Google scored a 77%, it’s lowest score ever.

This erosion of customer satisfaction may be leading to a drop in advertising ROI. According to a recent report from Analytic Partners, the return on investment from paid search dropped 27% from 2010 to 2016. Search wasn’t alone. All digital ROI seems to be in decline. Analytic’s VP of Marketing, Joe LaSala, predicts that ROI from digital will continue to decline until it converges with ROI from traditional media.

In April of this year, Forbes ran an article asking the question: “Is Google’s Search Quality Starting to Decline?” Contributors to this decline, according to the article, included the introduction of rich snippets and featured news, including popularity as a ranking factor and ongoing black hat SEO manipulation.

But the biggest factor in the drop of Google’s perceived quality was actually in the perception itself. As the Forbes article’s author, Jayson DeMers, stated;

It’s important to realize just how sophisticated Google is, and how far it’s come from its early stages, as well as the impossibility of having a “perfect” search platform. Humans are flawed creatures, and our actions are what are dictating the shape of search.

Google is almost 20 years old. The domain Google.com was registered on September 15, 1997. Given that 20 years is an eternity in internet years, it’s actually amazing that it’s stood up as well as it has for the past two decades. Whether Google’s naysayers care to admit it or not, that’s due to Google’s almost religious devotion to the quality of their search results. That devotion extends to advertising. The balance between user experience and monetization has always been one that Google has paid a lot of attention too.

But it’s not the presence of ads that has led to this perceived decline of quality. It’s a change in our expectations of what a search experience should be. I would argue that for any given search, using objective measures of result relevance, the results Google shows today are far more relevant than the results they showed in 2008, the year it got it’s highest customer satisfaction score (86%). Since then, Google has made great strides in deciphering user intent and providing a results page that’s a good match for that intent. Sometimes it will get it wrong, but when it gets it right, it puts together a page that’s a huge improvement over the vanilla, one size fits all results page of 2008.

The biggest thing that’s changed in the past 10 years is the context from which we’re launching those searches. In 2008, it was almost always the desktop. But today, chances are we’re searching from a mobile device – or our car – or our home through Amazon Echo. This has changed our expectations of search. We are task focused, rather than “browsing” for information. This creates an entirely different mental framework within which we receive the results. We apply a new yardstick of acceptable relevance. Here, we’re not looking for a list of 20 possible answers – we’re looking for one answer. And it had better be the right one. Context based search must be hyper-relevant.

Compounding this trend is the increasing number of circumstances where search is going “under the hood” – something I’ve been forecasting for a long time now. For example, if you use Siri to launch a search through your CarPlay connected device when you’re driving, the results are actually coming from Bing but they’re stripped of the context of the Bing search results page. Here, the presentation of search results is just one step in a multi-step task flow. It’s important that the result that is on top is the one you’re probably looking for.

Unfortunately for Google – and the other search providers – this expectation stays in place even when the context shifts. When we launch a search from our desktop, we are increasingly intolerant of results that are even a little off base from our intent. Ads become the most easily identified culprit. A results set that would have seemed almost frighteningly prescient even a few years ago now seems sub par. Google has come a long way in the past 20 years but it’s still losing ground to our expectations.

 

 

GE: A Different Company for a Different World

One week ago today, John Flannery took over as the new CEO of General Electric. He’s only the 3rd person in the past 36 years to have held the role. He takes over from Jeff Immelt, who in turn inherited the post from the iconic Jack Welch in 2001. Welch started his reign in 1981.

GE has been around for a long time. It actually predates the Dow Jones Index by 4 years (having been founded in 1892) and is the only one of the 12 original companies listed that still exists. It – perhaps more than any other company – serves as a case study for the evolution of the multi-national mega corporation. But GE is in trouble. Share prices are down. It’s struggling to retain its considerable grip on the industries in which it competes. Flannery has his hands full.

The GE story is also interesting because Jack Welch was the first rock star CEO. In 1981, when the Welch reign began, we were still very much in the era where sheer bulk equaled success. Size bestowed a considerable advantage on companies like GE. Welch recognized this and introduced the now famous “Number One or Number Two” strategy; where he pared down GE’s portfolio to just the industries where they could be either first or second in the world.

Ironically, given that he was lionized as one of the great corporate strategists of his era, Welch was relatively unimpressed with the classic interpretation of strategy.

“Forget the scenario planning, yearlong studies, and 100-plus page reports that “gurus” suggest. They’re time consuming and expensive, and you just don’t need them. In real life, strategy is very straightforward. You pick a general direction and implement like hell.”

It was this embracing of flexibility in planning that eventually led Welch to rethink the rigidity of his “One or Two” dictum. Jeff Immelt followed the same playbook, shutting down portfolios like finance and placing a heavy bet on high tech infrastructure. But despite Immelt’s best efforts, GE’s market cap dramatically eroded, shedding almost 30% and $150 billion in value over his 16-year stint as CEO. When you stack it up against Welch’s numbers – a 2790% increase in market cap in the 20 years his hand was on the steering wheel – it’s hard not to come to the conclusion that Immelt was a horrible CEO and Welch was a super star. But as logical as this seems, it’s based on faulty logic – what Phil Rosenzweig calls the Halo Effect. That fact was, the world of Immelt’s GE was a vastly different place than was the world of Welch’s GE, even setting aside mega events like 9/11 and the financial meltdown of 2008.

In those 16 years, many of economist Ronald Coase’s original reasons why a corporation exists disappeared. Most of them had to do with the market friction that came from a rapidly expanding physical market place. If you want the exhaustive analysis of this, go ahead and plow your way through the 600 plus pages of Alfred Chandler’s seminal work – The Visible Hand. But to pare that down to the barest essentials: it was much more efficient to actually build physical things and distribute them to a geographically dispersed market when you had a vertically integrated corporation where you could manage every step of the process. This was the world in which Jack Welch became the CEO of GE.

That’s not the world we live in today. Because transactional friction has been ruthlessly eliminated by technology, the efficiencies of the open market usually equal and sometimes exceed that of a corporation. Need a massive international transactional platform? The emerging blockchain commons can provide that. Need marketing capabilities that weren’t even dreamt of by even the biggest multinationals just a decade ago? Take your pick of almost 5000 MarTech vendors. Your start up office grown too big for your garage? You can even rent a corporate headquarters, complete with all the bells and whistles.

So the biggest question facing Flannery in 2017 is this: Are mega corporations – and, by extension, GE – even relevant any more? If we stick to Coase’s strict definition, the answer is probably no. But perhaps there’s another reason for corporations to exist: the critical mass of innovation.

Although she was vilified for it, I believe Marissa Mayer was on to this when she herded all of Yahoo’s teleworkers into the same physical location. In the infamous memo, Yahoo’s HR Director, Jackie Reses, said,

“Some of the best decisions and insights come from hallway and cafeteria discussions, meeting new people, and impromptu team meetings. Speed and quality are often sacrificed when we work from home. We need to be one Yahoo!, and that starts with physically being together.”

Mayer defended her decision by first acknowledging that “people are more productive when they’re alone,” and then stressed “but they’re more collaborative and innovative when they’re together.

Researchers have found that when the population of a city grows, the amount of productivity scales supralinearly. If you double the population of a city, you don’t just get a 100% boost in productivity, you also get a 30% bonus. Cities are the most effective innovation engines ever devised. And the reason is simple. When you pack a bunch of intellectually diverse people into the same space, magic happens. Mayer understood this. Unfortunately, the realization was “too little, too late” to save Yahoo but that doesn’t mean her logic was faulty.

As John Flannery steps into the CEO role, he may run full speed into the realization that the benefits once bestowed by being massive have turned into liabilities. What once made GE great now threatens to drown it. But there are still 300,000 different minds that work for GE. Frankly I’m not sure mega corporations can ever be relevant again in a friction-free marketplace, but if they can, the answer lies in the innovation potential of those minds.

Is Busy the New Alpha?

Imagine you’ve just been introduced into a new social situation. Your brain immediately starts creating a social hierarchy. That’s what we do. We try to identify the power players. The process by which we do this is interesting. The first thing we do is look for obvious cues. In a new job, that would be titles and positions. Then, the process becomes very Bayesian – we form a base understanding of the hierarchy almost immediately and then constantly update it as we gain more knowledge. We watch power struggles and update our hierarchy based on the winners and losers. We start assigning values to the people in this particular social network and; more importantly, start assessing our place in the network and our odds for ascending in the hierarchy.

All of that probably makes sense to you as you read it. There’s nothing really earth shaking or counter intuitive. But what is interesting is that the cues we use to assign standings are context dependent. They can also change over time. What’s more, they can vary from person to person or generation to generation.

In other words, like most things, our understanding of social hierarchy is in the midst of disruption.

An understanding of hierarchy appears to be hardwired into us. A recent study found that humans can determine social standing and the accumulation of power pretty much as soon as they can walk. Toddlers as young as 17 months could identify the alphas in a group. One of the authors of the study, University of Washington psychology professor Jessica Sommerville , said that even the very young can “see that someone who is more dominant gets more stuff.” That certainly squares with our understanding of how the world works. “More stuff” has been how we’ve determined social status for hundreds of years. In sociology, it’s called conspicuous consumption, a term coined by sociologist Thorstein Veblen. And it’s a signaling strategy that evolved in humans over our recorded history. The more stuff we had, and the less we had to do to get that stuff, the more status we had. Just over a hundred years ago, Veblen called this the Leisure Class.

But today that appears to be changing. A recent study seems to indicate that we now associate busyness with status. Here, it’s time – not stuff – that is the scarce commodity. Social status signaling is more apt to involve complaining about how we never go on a vacation than about our “summer on the continent”.

At least, this seems to be true in the U.S. The researchers also ran their study in Italy and there the situation was reversed. Italians still love their lives of leisure. The U.S. is the only developed country in the world without a single legally required paid vacation day or holiday. In Italy, every employee is entitled to at least 32 paid days off per year.

In our world of marketing – which is acutely aware of social signaling – this could create some interesting shifts in messaging. I think we’re already seeing this. Campaigns aimed at busy people seem to equate scarcity of time with success. The one thing missing in all this social scrambling – whether it be conspicuous consumption or working yourself to death – might be happiness. Last year a study out of the University of British Columbia found a strong link between those who value their time more than money and happiness.

Maybe those Italians are on to something.