Find terrorists, forecast sales: A game of predictive analytics


Last week I spent some time with LeapFrog Solutions, a Washington D.C.-based think tank specializing in federal agency communication and change management. Next to me in a conference room filled on one side with glowing windows and the other with whiteboards was Bob Derby, LeapFrog VP of Strategic Communications. On the phone was John Verrico, president of the National Association of Government Communicators and chief of media relations for the U.S. Department of Homeland Security’s Science & Technology Directorate. Bob is one of those management consultants with nearly 30 years experience who helps organizations plan their entire staffing and missions, and John is a rare data-communication guru hybrid who has helped federal agencies use analytics to track down terrorists.

Our mission was to help draft an upcoming seminar on Predictive Analytics, or more precisely, how to help enormous federal agencies take best practices from Corporate America to turn “big data” into actionable steps to achieve program outcomes. Frankly, it’s a head-scratcher.

The topic is enormously vast—predictive analytics range from modeling credit scores and terrorist threats to machine learning, data mining, and marketing campaign forecasts. Modeling varies across industries, from pharmaceutical giants predicting drug adoption curves to insurance companies evaluating the future impact of global warming on policies for homes on the Gulf Coast. How in the world do you boil the vast data flowing around every organization into a system for valid forecasts?

Yet as we drew diagrams around the topic, a simple model for predicting outcomes from complexity started to emerge.

When New York City’s Citicorp building almost fell down

First: What are predictive analytics? 

In one of our favorite data stories, documented by Joel Morgenstern in The New Yorker, in 1978 William LeMessurier, lead designer of the Citicorp Center skyscraper in New York City, received a call from a young engineering student who had been tasked with evaluating the Citicorp building, what was then the seventh-tallest building in the world. The student said that he had modeled wind forces on the building and thought the columns at the bottom—placed in the center of each building side, allowing for the bottom corners of the building to jut out over space several stories in the air—had been put in the wrong spots. The building, the student thought, might fall down in high winds.

LeMessurier chuckled at the kid’s naivety, got off the phone, but later re-ran some numbers and found, to his horror, the building actually might buckle in hurricane-force winds if they were what sailors called “quartering winds,” coming from an angle and hitting two sides of the building at once. Hurricanes did hit Manhattan every 90 years or so. And thus began a secret race to reinforce the Citicorp building’s structure from the inside out, all due to an error in math.

The Citicorp story is an illustration of predictive analytics: You are trying to build something (here, a 59-story tower that won’t fall down), need to evaluate how internal systems you control (steel beams) support the outcome, but also need to forecast external factors (high winds) that put stresses on your system. Your data must follow a chain of logic from outside to inside, prediction to event to result. If you model it right, you can control the outcome.

A simplified ‘gameboard’ for prediction

During our call, Bob and I started doodling in notebooks and the whiteboard, and a lucid model for “Predictive Analytics” emerged. The first draft looked like this:

pred analytics sketch

It’s extremely simple, really. The Y vertical axis, at left, shows two major areas—the external environment of things you cannot control, and the internal systems that you can control. The X horizontal axis, at bottom, shows time in three groupings: predictions, the period of time before your campaign or activity when you need to anticipate outcomes; events, the things that happen as your campaign is running; and response, how you react (in marketingspeak, this is often called the “optimization” period of your campaign).

Thus, on one side of the board, the things you can and cannot control; on the other, your forecasts, events and response.

Now, within the board, analytics and systems are grouped into six areas:

1. Game. This is where you make predictions for external environmental factors beyond your control, but which if gamed out, could be anticipated. (We use “game,” from “game scenarios.”) Example, a contender for president could have gamed out that a populist billionaire such as Donald Trump might enter the race and springboard off the undercurrent of economic dissatisfaction and fear in this country. No politicians can control Donald Trump, of course, but his current success in the polls could have been predicted. Marketers can game out competitor moves. Business leaders can explore Michael Porter’s 5 Forces models of competitors, suppliers, customers, market entrants, and market substitutes. Government agencies can game out scenarios of news events that might raise or lower public opinions of their missions. The future is uncertain, but major environmental influences on campaigns can be pre-conceived.

2. Forecast. This is where you model the variables you can control, and forecast outcomes using various statistical methods. In marketing, this typically involves comparing financial inputs into a campaign into forecasts for impact in awareness, response, conversions, acquisitions, and ROI. Forecasts can be tied to benchmarks of prior similar campaigns (overall) and specific communication pathway performance, channel by channel, medium by medium. A $1 million investment, providing no shocking “game” influences from step 1 above, should lead to XX,XXX results flowing out.

3. Test. This is real-time event management by testing variables you can control. Messages, ad creative, media channels, influencer outreach, conversion pathways, all should be tested with different flavors, colors, images, media tactics, and path structures. Even the very audience you are trying to reach should be tested. A classic example comes from automotive; when Honda launched its boxy Element mini-SUV in 2001, it originally thought buyers would be young men in their 20s who wanted a cool beach-surfing-camping vehicle. Buyers turned out to be dads age 35+, who wanted a fun small SUV to carry kids around in without looking like a mom in a minivan. As test data flowed in, Honda rapidly pivoted its ads away from pictures of the Element on the beach with fold-down seats for fully reclining (wink, wink, young men) to more family-focused advertising creative.

honda element ad

4. Monitor. This step means setting up rapid-response monitoring systems, so you can react to the world’s events as they happen. Tied closely to Step 1 above, gamemonitoring systems may include ongoing analysis of competitor organization communications campaigns; tracking of news stories; or sentiment monitoring of consumer discussions about your mission on social media using tools such as Radian6. One great, simple approach to setting this up is to draw a “touchmap” of data flows from all major outside factors (audiences, organization outlets that touch audiences, competitors, sales systems, news events, market entrants) to your internal data systems. Then, draw little stop signs where you have gaps from data you might need to your internal flows. Then, fix the gaps.

5. Measure. This is a simple word for a vast construction, and we’ve written elsewhere detailed guides to campaign measurement methodologies. But at its core, measurement means evaluating your campaign from the audience perspective—how their attention is reached efficiently; how their awareness is increased; positive or negative shifts in sentiment, responses, conversions; and the economic cost per desired action. Measurement puts the data dictionary and data flows against your Step 4, Monitor. It is worth being clear here that measurement does not necessarily mean investing heavily in new technology systems. For most organizations we work with, clients have all the systems they need in place already; instead of more investment in data systems or software, what is often needed is simple counsel in connecting the dots.

The danger of measurement is the output can be overly complex, leading to dashboards that look like a mathematician lost his breakfast. We recommend setting up KPIs (key performance indicators) that use a tree-branch structure, similar to drawing your family genealogy. What 2 or 3 main factors are you really trying to evaluate (such as your 2 parents)? For a marketing campaign, these might be lift in intent to use our product (a core brand metric) and cost per customer acquisition (a core direct-response metric). Behind them, what supporting data elements lead to these KPIs (similar to your grandparents and great-grandparents)? Etc. By nesting your measurement findings into a hierarchical structure that leads to a few core outcomes, you can both measure real progress in the major terms and also explore the more minor variables that create the chain to these results.

6. React. This is the punchline, the moment when you react to how the market around you has moved. But instead of reaction being “reactionary,” if you’ve successfully staged the preceding steps, you will be able to react smoothly and calmly to redirect your campaign as outside forces and audience results emerge. If the stock market crashes tamping down demand for your product gizmo, you will have anticipated this. If breaking news tosses a PR crisis your way that damages your brand, you will have a plan B and subsequent plan C in place. And whatever breaks, you’ve pre-installed measurement systems to gather news in as quickly as possible and gamed out rapid response pathways to maximize your influence.

It’s a very simple gameboard, filled with, yes, lots of complex work. But this work does not have to be expensive. We recommend as you consider predictive analytics that, instead of investing in a million-dollar data system, you throw this model on a whiteboard in your office, break out some yellow Post-It notes, and see how simply you can cover all the bases.

Update: We’ll be speaking on Predictive Analytics for the Future Nov. 5 in Washington, D.C. We’ll post details on the event in this space soon. If you would like to attend this morning session on Predictive Analytics, email




The pigeon-guided missile (building a marketing offer)

Screen Shot 2015-07-12 at 8.55.36 AM

One of the most important elements of any advertising campaign is not the brand or media channels or call center or landing page. It’s the offer. The typical U.S. consumer is bombarded with more than 6,600 TV ads and 375 minutes of radio commercials each month, so no matter how compelling your brand story, an offer is required to break through. So how do you construct one?

Our favorite models come from the work of psychologist B.F. Skinner and behavioral economist Richard Thaler. Let’s start with the story of Skinner’s pigeon-guided missile.

Birds in the front of a WWII missile?

You likely remember Skinner from a college Psych 101 class as the Harvard psychologist focused on human behavior, famous for the concept that dogs, if fed after a bell rings calling them to dinner, would start to salivate every time they heard the bell.  But did you know in World War II he trained pigeons to guide surface-to-surface missiles?

Skinner believed that all behavior was the result of such triggers, and complex actions were simply caused by the “chaining” of a series of trigger events. In the 1940s, he used these ideas to solve a vexing issue for the U.S. military: Missiles were expensive, but no autonomous guidance systems had yet been invented to fine-tune their direction as they flew through the air toward enemy boats once they had been fired. Skinner had the brilliant idea to set up three compartments at the front of a missile with lenses pointed in three directions, projected on three internal screens. A pigeon would be placed in each compartment, trained to peck on a screen if it saw the image of a battleship, and the peck would guide the missile to turn in that direction. Skinner received $25,000 in funding — about $420,000 in today’s dollars — but the military eventually stopped the project out of, er, disbelief.

To put a finer point on it, Skinner conceived there were two forms of behavioral triggers:

Respondent behavior, where an outside event makes a being respond. Pigeon sees image of battleship, and used to getting grain, it pecks. You smell toasted food from the oven, and you get a pang of hungry desire.

Operant behavior, a more complex trigger, in which a person performs an action started internally that is then reinforced or discouraged from the outside. Say, an executive smiles brightly in a meeting one day and everyone suddenly warms to her idea; she learns to smile repeatedly when she wishes to influence others.

Everything animals or people do, he suggested, came down to chains of respondent (outside) and operant (internal) triggers.

The components of a good offer

What’s interesting about this (rather crazy) missile idea is Skinner, when tasked with setting up the fastest response mechanism possible, guiding a missile, resorted to respondent behavior triggers from the outside — the simple visual cues that made pigeons peck. There were no complex chains of multiple trigger events. When a response was needed fast, he relied on simple triggers first.

Advertising offers work in a similar vein. While internal marketing departments tend to think deeply on their products (as is their job) and fall in love with the nuance of how they may be slightly different than competitors, outside customers who rarely think of you must make a decision in split seconds as to whether your story is enticing. The external trigger needs to be lightning fast — akin to the smell of warm toast, or an image they immediately recognize.  

Richard Thaler, the University of Chicago mind who invented the field of behavioral economics in his landmark 1979 paper “Toward a Positive Theory of Consumer Choice,”  created a model for building rapid-response triggers. In essence, Thaler suggested humans are irrational, busy, and apt to emotional judgments. Fairness is a classic cloud that causes people to make illogical decisions. Logically, if a business deal could earn you $10,000 in profit you should take it. But if you felt the deal should be worth $20k instead, you might walk away because the $10k offer feels “unfair.” On the flip side, if you thought $5k would be fair, you’d rush in to close the $10k deal, because you’ve somehow won extra. Thaler’s research found that the actual end result isn’t what matters for humans as much as the perception of whether we are victorious along the way.

Thaler suggests the best way to influence humans is to separate gains, minimize losses, and use a reference point to play to the human psychology that we all like to win more than we lose.

Combining Skinner with Thaler, here are the components of a good marketing offer:

  • Set a reference point to “frame” your value. Decisions on gains are often made compared to a fictional “reference point.” Thaler found most people are bad at judging value, so we all like to be told what the starting reference point is on a product or service. A man may not want a suit priced at $500 … but if the suit is on sale marked down from the (fictional) price of $1,000, he suddenly covets it, now “50% off.” So one way to influence people is to show how the cost of your product is far below a (fictional) higher price.
  • Set a time limit suggesting they’ll lose something. Thaler found that most people hate losing $100 more than gaining $100. We have a steep aversion to loss, because psychological it seems unfair. So set up a fictional loss, such as a time-constraint on an offer after which “savings” will disappear, to spur behavior.
  • Segregate gains suggesting they’ll win more. In several studies, Thaler found that if given a choice between winning a lottery two times for $50 and then $25 respectively, or winning just once for $75, the majority of people preferred to win twice. Again, numerically this makes no sense, but emotionally we feel better about the “winning events” than the actual value. So marketers can separate gains into distinctive categories — “get X savings and also Y free!” — for the same value to drive greater response.
  • Integrate and minimize losses. Unlike the lottery example above where people prefer to win smaller prizes more often than one big prize, if given a choice in pain, humans prefer to take a loss all at once. So make the actual cost (pain) of the product as simple and minimal as possible. Don’t charge for the coffee and the coffee cup and the lid and the sleeve separately, because even if they all add up to $1.50, people will be turned off by the multiple cost pain points. Don’t charge for the product and installation and service separately. Instead, lump all costs together, and if possible spread them out into low payments to make the “pain threshold of price” as low as possible. The best price point is simple, low, and nearly invisible. This is why Netflix charges one low fee per month for streaming video, and not a separate fee for every movie you want. This is also why has had huge victory with its “free shipping” that really comes bundled into an annual Prime membership fee. The loss, or price paid, comes once, but the benefit feels free every time you order something throughout the year.
  • Consider bundling or price obfuscation. All of the above can be combined to maximize gain segregation (the series of good things the consumer “wins”) and minimize loss (the financial cost the consumer “loses”) with bundling. Omaha Steaks is the classic example, where a package of bacon-wrapped filet steaks comes bundled with burgers, franks, French fries, vegetables and something called apple tartlets. All marked down from $173.00 to $79.99. Is this a good deal? We don’t know. But this complex bundle coupled with a simple price below the reference point sure sounds desirable.
  • Avoid complex service claims and focus on desire. Your long-term customer service may be incredible, and that’s great for building loyalty, but the “offer” in your ad must be a simple understandable hook that gets the consumer pigeon to peck. New response triggers must come from the outside and be easily understandable — an element of your product or service that is immediately desirable, and that consumers must feel if they don’t act now, they will lose.

In sum, marketing offers must be weighted toward Skinner’s “respondent behavior” — fast triggers from the outside that are immediately understandable — vs. “operant behavior,” the other more complex triggers learned over time. There is a difference between the elements of your product/service that build long-term loyalty and the sizzle that gets consumers to respond right now.

Like the enemy battleship screened before the pigeon, you have to make your offer immediately recognizable. Without a strong offer, you’ll miss the boat.

Twitter is so hungry for new users, it’s rebooting

twitter bad ads


Twitter is about to launch a major redesign of its central flowing cascade of tweets, code-named Project Lightning. Whether you call this an attempt to make live events an add-on (Mashable) or an attempt to kill Twitter to save Twitter (Wired), it is a huge shift. Soon, instead of relying just on friends to tell you what they are thinking or linking to, you can click a central button on Twitter to see what a bunch of … editors think you should see. Videos. Photos. News. All curated by experts and algorithms to help you easily learn about current events.

Twitter is about to become the Walter Cronkite of social media, an AP Wire feed of the world’s top topical buzz.

Why would Twitter make such a change? Basically, Twitter is in deep doodoo with Wall Street, despite growth that would delight most businesses. Quarterly revenue has soared from $18 million in Q1 2011 to $436 million just four years later … but investors are worried about user growth, not just dollars. In a shrinking social media bubble where startup valuation is still tied to (supposed) network effects of interconnected eyeballs, Twitter’s annual user growth has declined from 50% in 2012 to 18% in 2015 and is projected to dip below 10% soon. Twitter’s stock is down 50% from its high two years ago; investors punched it down 20% alone in April when missed earnings were announced; CEO Dick Costolo is so fed up with the pressure, he’s quit.

While Twitter now has 302 million monthly active users, that number is way below former company forecasts, which back in 2009 estimated Twitter would top 1 billion users by late 2013. Twitter missed that mark; Facebook cleared it with 1.4 billion users; and today, Twitter seems hopelessly lost behind its older brother Zuckerberg. 

7 ways Twitter can survive

Investors are worried because Twitter, in their view, must generate money; money is tied to eyeballs; and those eyeballs are growing sleepy. What can Twitter do to grow?

1. Get users. The first way is to grow its user base by making it easier for people to sign up, and stick, with Twitter.

2. Boost ad rates. The second way to boost revenue is to increase ad rates, difficult given the competition from Facebook’s CPC model and Google search, both of which perform extremely well in driving low cost response.

3. Increase ad volume. The third way is to increase the number of ad units within the stream — also problematic, because higher noise levels in social networks spur customer churn.

4. Sell media space outside Twitter. The fourth way would be to increase, somehow, the number of ads sold outside the Twitter ecosystem — quite possible, given Twitter’s April 2015 acquisition of TellApart, a tech firm specializing in cross-device and cross-platform ad retargeting. Of course, this would have to be supported by good data on Twitter users, and unfortunately Twitter has little of that. Facebook was smart enough to ask you about all your interests and get you to Like hundreds of brands … but Twitter, shy in the corner, forgot to ask.

5. Sell something else. The fifth path would be to expand to ancillary revenue streams, such as selling data on users (which Twitter has little of), taking a cut of e-commerce or Twitter-inspired offline transactions, or getting into actual mobile payments.

6. Paywall. The six option would be to charge a subscription for usage by consumers. Ha. In a world where information wants to be free, a Twitter consumer paywall would be death.

7. Charge media partners fees for access. The seventh would be charging other media platforms a fee for integration with Twitter promotions, sort of an extension of its current ad model.

Hm. That seems to be about it.

Twitter can only squeeze you so hard. So it wants more yous.

Most of our proposed solutions above — Nos. 2 through 7 — require Twitter to clearly push more ads or extract more data or money from current users. That’s a dangerous game, because boosting clutter in any network can kill it. (See: email marketing and telesales.) We are already seeing signs Twitter is at capacity in pushing ads on users. When the network missed revenue targets in April, analysts said that ad revenue per user was actually up (meaning each of you had more ads in your stream) while ad performance was down (new direct-response ad units didn’t perform to standards, meaning fewer of you clicked on Twitter ads). Too many ads with too little response suggests … Twitter is wringing its current users dry with too much marketing.

So solution No. 1 above — finding new users — is the best path forward. Here, Twitter’s challenge is it has become very difficult to use as more and more features are added.

Enter Project Lightning, with an easier learning curve

The service has always been complex with a steep learning curve. The Twitter UX now requires following people, hoping they follow you back, contributing to a tweet stream, a personal profile page parsed into tweets, tweets and replies, or photos and videos, lists, favorites, direct messages, and a core communication complexity involving short text but potential photo, video, URL links, @’s, notifications, and hashtag additions, all within a character limit. Someone new to the Twitter party may now find the once-simple text service so complex they think, WTF?

But Twitter’s Project Lightning circumvents all that learning. Soon, if you join Twitter, you won’t even need to connect with anyone to dive in. You can just click a big button in the center of your screen or mobile device and get a curated list of cool news/videos/images tied to the major hot topics of the day. Obama sings at a service. A sports team wins a championship. A hot movie star exposed on the red carpet. You’ll immediately get the warm buzz of real humans sharing these events, a verisimilitude of personal connection, all concocted by the algorithmic robots at Twitter’s HQ.

It’s a smart idea, making it easier for new users to join. But this experience will be a very different Twitter — sort of a virtual curated network of no friends inside a virtual network of somewhat fake friends. A news circle in a friend circle, growing the circle of users to attract the bigger circle of marketers.

Good luck with the new button, Twitter. We know your investors are counting on it. We just hope your constant expansion into the ancillary media world to attract less-sophisticated users doesn’t kill the fun we had trying to make 140 simple characters work.


When native advertising works: ‘The Frame’

Frank and Claire

Two years ago we wrote a column in Digiday expressing dismay over the rise of “native advertising,” the brand-sponsored content in which marketers themselves produce editorial material. Almost all publishers demarcate native ads with labels, saying “sponsored ad,” but this isn’t enough. Our beef then was that native ads often misdirect users by disguising the source of a message — after all, a good “native” piece looks like a real quality piece of content, yet its underlying mission is to promote a brand or sell a product. Even if the source is disclosed, the underlying misdirection remains. Native advertising is too often the publishing equivalent of that old college chum who shows up at your reunion party only to push a business card.

Or is it? Recently we’ve seen an evolution in quality of native ads that has made us rethink this objection. We call this type of positive native advertising “The Frame.”

The Frame: Microsoft’s Modern Workplace

Here’s a great example of The Frame. Today we stumbled upon a video interview with Kevin Ashton, who in the 1990s was a junior manager for P&G when he noticed a problem. Retail stores kept running out of certain products that would surge in popularity among consumers, but his inventory tracking systems were always one step behind. Keeping up with this inventory game was like whack-a-mole, where a new hole would appear in store shelves just when he filled another. Ashton realized he needed a new way to track the location of products where they were in time, and so invented the idea of putting little RFID radio chips in each product.

“We could then sense where things were by themselves,” Ashton says, and he put his idea into an internal P&G PowerPoint called The Internet of Things. Today, Ashton’s coined phrase is the hottest idea in all of technology as everything from couches to refrigerators begin talking to each other.

We learned this story on the birth of smart devices not from Wired magazine or IEEE, but from a sponsored video by Microsoft, as part of its “Modern Workplace” series. Microsoft has created a PBS-style video documentary series on major issues facing business, with interviews of real-world luminaries such as Dean Kamen, inventor of the Segway and emerging robotic artificial limbs. It’s native advertising, but done in a way that provides immense value to readers.

The Atlantic’s Scientology Misdirection

We’ve written before that there are only three times of native advertising: “The Frame,” the most innocuous, where a brand sponsors text or video but does not insert itself into the picture; “The Insertion,” where the brand itself is pushed inside the story, such as a piece on energy with a case study by Chevron, all paid for by Chevron; and “The Misdirection,” in which a publishing platform runs something where the paid source attempts to mimic unpaid material, misdirecting the audience.

The Misdirection is where brands really get into trouble. When The Atlantic ran a glowing Scientology article online in January 2013, the piece was clearly marked as Sponsored Content, and used different color headline fonts than the publication’s main editorial. But readers screamed, confused as to why a leading authority in journalism would apparently publish a puff piece about a controversial religion. Technically, The Atlantic did everything right: disclosed this was an ad, ran it is slightly modified content. But the audience rebelled.

A day later, The Atlantic editorial team published this apology:

We screwed up. It shouldn’t have taken a wave of constructive criticism — but it has — to alert us that we’ve made a mistake, possibly several mistakes. We now realize that as we explored new forms of digital advertising, we failed to update the policies that must govern the decisions we make along the way. It’s safe to say that we are thinking a lot more about these policies after running this ad than we did beforehand. In the meantime, we have decided to withdraw the ad until we figure all of this out.

Then, The Atlantic got native right 

Apparently, The Atlantic did figure it out. In March of this year it rebounded with a quality native advertising feat, a deep political essay on the dynamics between power couples in the presidency, leading with the question of what Bill Clinton will do if Hillary is elected. The sponsored piece promoted Netflix’s House of Cards Season 3, in which Frank and Claire duel for power, but only by carefully framing the bit, not by pushing the popular series’ characters too deeply into the content. Digiday called us for comment on the ad and all we could say sincerely was, well done. The piece tapped a real news issue — what’s up with Bill and Hillary? — the deeper interest of marital relationships and power, and promoted House of Cards oh so gently, all clearly labeled. We read the entire section twice.

So what is a marketer to do? The greatest test of native advertising may be to answer two questions: is the source of the material truly apparent (your brand is publishing this, so say so clearly) and does the content provide real contextual value (beyond just shoving your brand in the reader’s or viewer’s face)?

What Robert Scoble said

Way back in 2008, we interviewed Robert Scoble for his opinion over the rise of sponsored posts in blogging. The debate then was whether bloggers should take money from brands, and also should bloggers disclose they had before they promote a product. We raised the question after a top blogger, now a friend, accepted a $500 gift card from a major retailer around the holidays and then sprayed his readers with a profile on how great shopping at that store was; even though he disclosed he was paid, the event left a sour taste in our mouth. Thanks a lot, we thought, for sharing how great that chain is as you spend its money to buy yourself stuff. Scoble thought on this and responded that in quality native advertising, not only should brands and publishers reveal the source of the material, but the material must come off as authentic.

“The brands that protect their credibility and authenticity go up and the ones that don’t go down,” Scoble told us. “This world moves so fast, if you get caught selling out your readers, you will get exposed and derided, and you’ll be less for it.”

Not every brand can pull off native at high levels of editorial quality and ethical integrity. It costs money, and you may need products or services that are closely tied to real human needs and societal concerns so your story truly resonates. But then again, if your brand isn’t relevant to people and society in a way that provides true value, you have deeper problems than how to spin your advertising.

The dismal rise of smart TVs

Screen Shot 2015-02-22 at 5.36.09 PM


Why are smart TVs not scaling?

Nielsen published some interesting stats recently on technology adoption in the United States. Broadband access to the Internet is now near 80%, meaning even Grandma has it. Smartphones have skyrocketed from about no use in 2007 (when they were first released by Apple as a category) to adoption by 3 of 4 adults. Tablets went from zero in 2010 to 46% last year, and today should be in the hands of 1 in 2 consumers. But smart TVs are trailing … in 2014 reaching only 13% of the U.S. population. At that lackluster growth rate, in five years only 1 in 4 U.S. households will have one.

Smart TVs are basically large video screens connected to the Internet, allowing you to “stream” content online, from Kevin Spacey taking over the world in Netflix’s House of Cards to YouTube videos. There are numerous reasons adoption may be slow: the average U.S. household already has three TV sets; consumers recently went through mass spending on flat panels, as they emerged as a sexy category about a decade ago, so may be reluctant to upgrade yet again; and the remote controls of the smartest TVs still don’t lend themselves to typing in commands for Internet video searches. Between the cost outlay and the lousy absent keyboards, it’s little wonder few have adopted to Internet-connected flat panels.

But there is a deeper psychological issue at play, too. When Robert Sommer first wrote of “personal space” in 1969, he suggested we actually have three fields of taking in information: an intimate space near our face or ears, similar to a lover’s whisper; a personal workspace about arm’s length away, the distance of tools in our hands; and a social space from about 4 to 10 feet away. Today’s technology fits perfectly in each of these fields: mobile is intimate, laptops are personal/work space, and TVs are social. We are more likely to speak up in our intimate space (“Honey, please move your elbow”) and more focused on listening in our social space (“shh, don’t interrupt the storyteller.”) This is why our thumbs crawl over mobile smartphone keyboards but with TV, we just want to chillax for the show.

Don’t get us wrong. TV is still king of all media. Despite all the hoopla over digital and mobile, consumers spend more than 4 hours a day letting the blue light of cable bathe over them, outpacing time spent on any other communication devices. But our utility of television is one of social receptivity. We don’t want to engage with big screens, but instead, wish for them to entertain us without nuanced input. Like stories from around a campfire, the streams that come from TV are meant for us to be received as passive entertainment. Our guess is “smart TVs” may never take off, even as screen resolutions grow sharper and the flat-panels increase in size until they are as large as your basement wall. Our modality is simply passive as we watch Kevin Spacey. When we want to truly engage, we turn to the mobile Twitter interface in our hands.



Media predictions for the far-forward future

woman hologram

Put down your smartphone app and think far, far ahead. Media prognosticators rarely do this, perhaps because advertising clients and digi-journalists gain more from toying with the latest Twitter group chat update than they do by conceptualizing the state of media 100 years from now … but a far-forward forecast could be worth the effort.

So let’s play the prediction game.

Before we start, here is our inspiration: the brilliant book “The Next 100 Years,” in which George Friedman examines the macro trends of history in attempt to predict world events through the coming century. It’s an amazing feat, reeking of intellectual arrogance, to try to foresee 100 years of future events … until the reader discovers that Friedman has a solid methodology.

Friedman bases far-forecasts on geopolitics — the combination of national resources, locations on the globe, culture, and economics — which has ongoing patterns that make shocking events, such as World War II or the terrorist attacks of 9/11, predictable. Individual players on the planet, even presidents or kings, typically have far less power than we imagine, and must play upon a chess board that is already set. The future, it seems, can be predicted, if you really examine the macro trends. For instance, looking backward, Friedman argues:

  • It was inevitable Europe would become a global power in the 1800s, because it needed supplies from Asia, and after Turkey cut supply routes over land Europe sought a route west to India by sea and thus learned to manage the oceans — controlling global commerce.
  • It was inevitable that the United States would win the cold war over Russia in the late 1900s, because allies to the United States could “sell in” to its vast consumer demand set, making U.S. friends rich, while Russian allies might get weapons but end up impoverished. 

And, looking forward, he suggests:

  • The United States will remain the leading world power in the 21st century, because it controls the world’s oceans, due to its fortuitous placement between both the Pacific and Atlantic oceans, thus controlling trade.
  • There will undoubtably be another horrible global war in the 21st century, given the tensions between the rich and poor and the continued belief of humans in their personal nation states.
  • And this war, like all others, will eventually end, and generate new technology systems from the wartime investment that lead to sources of clean power and communications we today can barely imagine, such as microwave energy beamed down from outer space — the most efficient way to capture energy from the battery in the center of our solar system, the sun.

Oceans, human antagonism, and sunlight are all constants, and they will define our future.

So what is the real far-future of media?

Here are our predictions: (1) Environment monitoring, driven by sensors around us; (2) virtual visual overlays that are constantly on, created by the inevitable shrinking of screens until they fit in your contacts; (3) ambient personalization, as you control what you see everywhere; and (4) societal upheaval as we relearn how to interact with other humans in a virtual world.

1. Sensors everywhere — Behind all the Nielsen updates on multiscreen use or Pew reports on social media fads, the truth is information flows only two ways — to us or from us. While media writers remain fascinated with toys, the biggest trend in information flow is the spread of connected sensors in all devices. The iPhone 6 has six sensors built into it — including proximity, motion sensor/accelerometer, ambient light, moisture, compass, and a gyroscope — coupled with GPS features that pinpoint the phone’s location. The Disney Research lab has created a Touche interface that can turn the surface of any object, such as a table or couch, into an input sensor by monitoring vibrations created by human contact. Philips has launched “design probes” that explore tattoos with sensors that disappear based on touch, and clothing that changes color based on your mood.

With sensors everywhere, you will be tracked. Tracking will require control, so humans will use that to personalize their environments (a benefit) while suppressing unwanted third-party oversight (a cost).

When you walk into a room in 2070, the room will know who you are.

2. Screens everywhere — Concurrently, the spread of screens is obvious. At SXSW Interactive last spring, the head of the Consumer Electronics Association, Gary Shapiro, said that within 10 years consumers will buy wall TVs — or whatever we will call high-resolution digital screens that fill an entire wall. Apple has a patent for holographic wall screens that project 3D images to both eyes of each user in the room, without them wearing googles, by monitoring the location of their heads. And Microsoft has just launched a HoloLens goggle prototype that overlays 3D images on reality with a wider field of vision than the (recently aborted) Google Glass. As big TVs grow into walls, little visual screens will also shrink into contacts.

With screens everywhere, you will see whatever image you wish to pull up.

3. A personalized universe — The great media prediction for the next 100 years is that humans will be able to retreat into completely personalized bubbles of vision, overlaying data about others in their contact lens, porting their images into virtual meeting rooms thousands of miles away, and pulling entertainment into the real world around them. Because if everything (from couch to table) senses you, things will recognize your preferences, creating demand for automated content that overlays your reality to meet your unique needs.

From a content creation perspective, this will unlock a gold mine of opportunity for film (hologram) producers, game designers, social media entrants, work/office productivity software, pornographers (always the earliest refiners of visual technology), religions (where belief systems could now be “seen” as reality), and yes, marketers (who will find a way to support this content with some form of advertising over there on the side). This information rush will become fuel for economic growth, with visual services an entire new platform for monetization.

4. Societal unrest — These media trends are our predictions, not Friedman’s, but he has a point that may refine ours: Every evolution in society comes with unintended consequences. The vast rise of visual screens and the concurrent measurement of human personal preferences on every surface device may unearth new social dynamics we cannot anticipate. Will people become more gregarious as they seamlessly are able to beam their avatars into the world? Or will humans retreat into dream bubbles, like those poor enclosed battery souls in the Neo Matrix, asleep in cocoons while they envision a fantasy of greatness?

We cannot predict that. But one thing is certain: The far-forward future contains much more than an iPhone.

Why Google beats Facebook in mobile conversions

facebook mobile

Adweek seemed surprised this week to report two findings — a whopping share of Facebook and Google ads are now being served on mobile, and yet for some reason, both lag behind desktop ads in conversion rates (the % of people who click on an ad who end up completing the desired action, such as filling out a lead form). So let’s break this down:

1. First, both Google and Facebook have much lower conversion rates on mobile than on desktops. Marin Software monitored $6 billion in ads, or about 3 billion ad clicks, and found that while Facebook received 63% of all ad clicks on smartphones and tablets, only 34% of its conversions happened there. Google did a little better, with 39% of all paid search clicks being on mobile and 31% of all purchases made there.

2. Should mobile marketers panic? Well, no. Duh. Mobile devices have small screens and awkward touch keypads, so conversion will be lower, of course. Have you ever tried filling out a web lead form or typing in credit card information on an iPhone? So the overall trend will be for consumers to explore ad information, if interested, on mobile, but then convert on desktops (or even by telephone or physical store) later. This explains why Facebook mobile ads have only a 0.3% conversion rate vs. its desktop ads converting on average at 1.1%.

3. Now, within this race, why did Google still outperform Facebook on mobile conversions? Modality. Google search ads are triggered by consumers who instigate a search for a particular product, so they are already leaning toward conversion. If you punch in “airline tickets to Florida” on your iPhone, odds are you may be thinking of making a travel purchase. Facebook ads, instead, are pushed out to target consumers who have expressed no immediate interest in buying the product — so even if they click, their mode may be one of cursory exploration vs. immediate consumption.

All of this is to say that mobile ads can work very well in reaching audiences with information about a product; marketers should also take heart that most conversions happen subsequently across different channels. 100% of television ads, for instance, have conversions elsewhere — web, phone, retail store visit — because no one buys anything by clicking on a TV ad. (You can’t.) Imagine the histrionic Adweek headline: “U.S. marketers spend $70 billion annually on TV with a 0% response rate! Why aren’t there conversions?” Um, yeah.

Break out the regression analysis

The real solution to cross-channel mobile is to use multichannel measurement, evaluating the responses from a cumulative mix of digital or traditional advertising media. (We do this for clients with a mix of software and statistical regression analysis*; it’s quite fun.) The real story may be those early ads on Facebook spark interest that bring people in to Google search later, just as a TV campaign can build lift across physical stores. All ads are connected. The data trails between them are complex, but can be measured.

So keep making mobile impressions, marketers. Your spouse didn’t marry you after your first impression. This is not to say that first impressions don’t count.

* If you are new to stats, “regression analysis” sounds complicated but the concept is simple: It models relationships between variables, such as X television schedule and Y Google searches, to find with relative certainty how they are connected. (Imagine you go out partying and the next morning have a wicked headache. If you model this with enough parties over a year,  you could say with relative certainty: partying in fact does cause headaches.) Regression analysis is useful in evaluating how different, unconnected media tactics — and outside events, such as major winter storms or competitor behavior — work together to influence responses to ad campaigns. Without this type of analysis, you might make a mistake of shutting off one media channel, such as TV, which in reality could be driving thousands of customers in elsewhere in your marketing ecosystem.


Why marketers know if you’ve been naughty or nice

santa watching

There is a story about a jolly old elf who tracks your behavioral data carefully, spies on you even when you’re sleeping, and runs algorithms to assess whether your actions are more positive or negative than social norms. Based on his calculation, the elf will reward you with financial gain in the form of material goods or will deduct from your status by tricking you with what looks like material goods but in reality turns out to be lumps of coal. The system is extensive, including a database of every youth in the world, and is updated annually. If you don’t like this surveillance, good luck: The elf’s privacy policy is unpublished, the observational data cannot be accessed by individuals, and your only recourse to correct misinformation is to send handwritten postal mail to the elf’s address at the North Pole.

Perhaps these childhood stories are why people often freak out about data. The legends of people recording others’ actions, especially those of children, as a form of behavioral modification have been with us for millennium. In Bavaria, the Santa myth is actually split into two figures, a Saint Nicholas who rewards good children with gifts and a devilish, horned Krampus who punishes bad children. Japan has a similar tradition, with an Namahage figure played by men wearing huge, ugly masks, who knock on doors and warn children not to misbehave. Religion is filled with data tracking, starting with God watching Adam and Eve’s naughty apple-biting in Eden, moving on to the widespread but vague idea that somehow all of your actions in your lifetime are being observed for a final post-death judgment. In our deepest beliefs, we perceive there is a connection between what we do, how others record it, and how we will be rewarded.

Which brings us to marketing surveillance 

If you collect enough data to form a baseline for comparing people, you end up with a “database” — and this idea has been around for at least 400 years.  In America in the 1600s, clergy tracked births, marriages and deaths; officials called “tythingmen” would also enter homes to inspect families for observed moral behavior. The first consumer database in the United States was set up in Massachusetts in 1629 to track property ownership. As data expanded, intrusions did too. In the early 1700s, U.S. postal mail was opened regularly to spy on message content.

And then marketers figured out they could make money from all of this information. Database marketing started in the 1940s, first driven by direct-mail marketers (who needed target lists of consumers to mail things to and then calculations to see what worked), later by credit-card companies and banks (who rapidly learned that not all consumers have the same credit risk), and then in the 1990s by Internet marketers who realized they could measure a treasure trove of consumers’ online behavior. While the basic approaches are the same — identify potential customers, differentiate by their value to you and what they need from you, continue to gather more information through interactions, and then customize your response — the cycle time of data marketing increased. Direct mail list updates used to take months; if you purchased a pair of boots at a store in December, it might be March before another company’s boot catalog showed up in your mail. But the Internet enabled a cycle time of identification, differentiation, interaction and customization within days, hours, and now even seconds. Visit, look at shoes, don’t buy them, and you’ll see ads for similar shoes on other web sites within seconds. The prevalence of such digital “retargeting” has gotten so rapid that many consumers are beginning to freak out.

The systems are growing ever-more sophisticated. Digital media vendor Rocket Fuel has begun testing device fingerprinting to track consumers by their individual mobile phones; in a recent campaign for Brooks running shoes, it identified the mobile devices of everyone standing along the running route of the New York City marathon, and then later served ads to those devices for running equipment long after the crowds had dispersed to Baltimore, California or even foreign nations. Digital marketers can pick up the IP address of a home’s Wi-Fi connection, and then retarget multiple devices — based on a trigger of one person’s behavior — across the many iPhones, tablets and computers residing in a household. Creative-based retargeting is another digital approach in which banner ads or online videos can be retargeted based on a single ad appearing on any web page, whether or not a consumer clicks on it; for marketers, this provides the advantage of being able to “lift” a publisher’s audience, such as a reader of, and chase that individual around the web later with a pretty good idea of their demographic profile based on the original reading material.

Consumers are rebelling, so what is the balance?

Not everyone is happy about this. Early in 2014, a survey by Truste, a global data management company, found that 74% of Internet users had increasing worries about the use of online data. While only 38% expressed worry about government surveillance, 58% said they had concerns about business use of their personal information. Beyond simple consumer annoyance, the growing use of online data may actually be harming marketing results. 83% of survey respondents said they were less likely to click on an online ad due to privacy concerns. In a deeply ironic circle, the data collection sophistication used to make online marketing work better may actually be depressing response rates.

Smart marketers are recognizing this and beginning to tone down the creep-factor of retargeting, using tactics such as impression caps, dayparting, ad creative versioning, and opt-out options to allow Internet users more breathing room before they are inundated with braying offers.

Data tracking will not ago away, because it is how all of us assess the outside world to calibrate our actions. Marketing in particular is all about treating different customers differently, as the great Don Peppers once wrote — after all, if you have unique needs, you should receive messaging about products or ideas that appeal to your interests, and marketers who play this right will gain greater results from their advertising investments. Just as parents and Santa Claus watch over children to assess behavior, other people will always be watching you too. The practice isn’t creepy in and of itself; what has gotten scary is the instant cycle time it takes someone else to pass their judgment. For our clients, we recommend looking beyond just response and conversion rates to also assess the real end customer experience. You’re trying to share information that benefits the customer, so pace yourselves, people. Everyone likes an elf who brings presents, but we all get nervous if he’s watching us too much.

Understanding multiple device use: Meshing, shifting and stacking

woman smartphone 2

A young woman plops on the couch, turns on the TV, and as her favorite reality show casts a blue glow across the living room … she also boots up her smartphone to check on her friends in Facebook and … also swipes open her iPad to play Words with Friends.

All at the same time.

Marketers who want to reach a consumer on all devices concurrently often struggle with understanding how these touchpoints interconnect. The biggest challenge is consumers often use all these devices for different things. While TV is on for video entertainment, mobile devices are used more often for playing games or participating in social media.

Behind this is the fast-growing trend of, yes, people using mobile gadgets plus TV at exactly the same time. BI Intelligence just reported that 45% of all smartphone use by U.S. consumers age 16-44 is done with the TV on, as well as 37% of laptop use and 55% of tablet use. If mobile is ascendent, TV seems to be its constant companion.

To address this puzzle, we’ve searched for frameworks on how people actually use different screens at the same time — and found the best from Monique Leech, an analyst at global research firm Millward Brown. With a hat tip to Leech, here’s our own interpretation of her findings: there are three core ways people use multiple devices and each requires a unique marketing strategy.

Meshing: ‘Hey look, tennis is on ESPN. Let’s read too!’

“Meshing” is when people use two or more devices to watch directly related content. For instance, when Jane Smith was watching the Super Bowl on TV a few years ago, she was surprised by a blackout in the stadium lights, and turned to Twitter on her handset to chat about it. Oreo famously leaped on this moment by tweeting “You can still dunk in the dark,” and Jane would have laughed. Marketers who want to leverage “meshing” behavior can either target integrated advertising content, such as a buy on a weekend sports event and a concurrent media buy on, or deploy “real-time marketing” responses on social media during major awards shows or sports events.

Alas, meshing is only part of the story, and typically not the dominant form of concurrent media use. Putting an ad on to match a pro tennis tournament on TV at the same time may not always be the best approach. The next behavior, “stacking,” explains why.

Stacking: ‘Hey look, Walking Dead is on TV. But let’s chat on Facebook too!’

“Stacking” behavior is different, and more common, in which, say, James Smith is watching “The Walking Dead” on television while simultaneously chatting with buddies on Facebook via his iPhone. Stacking means adding different content from one device to unrelated content on another media device, all at the same time. Numerous studies show this is the dominant form of concurrent device usage. recently monitored 470 consumers for a month and found they spent an average of 3.3 hours on smartphones per day with the top activities being emailing, searching the Internet, or social networking. Tablet behavior was similar, with social networking and reading at top. What’s most interesting is so few reported watching TV-related content simultaneously on mobile gadgets, it didn’t make the list.

For marketers, this means you can’t just buy ads on to align with viewers watching CNN on TV. Instead, you must explore audience targeting across content platforms at simultaneous times, to reach consumers on Facebook or in a game while they watch a show on television.

Shifting: ‘This content is fun, but I’ll pause now and continue it later.’

The third form of multiple device usage, “shifting,” is one of shifting from one gadget to another while pursuing related content. This could be as direct as watching part of a Netflix movie on a tablet and finishing it on TV, or more nuanced such as researching a trip to Italy on a smartphone and then completing the reservation via a computer browser window.

This “shifting” device behavior poses two challenges for marketers, in targeting and measurement. For targeting, it requires understanding how different media touchpoints may be used in sequence for a consumer to learn about, explore, consider, and then consummate a desired action — and for measurement, it means the combined impact of all these channels must be evaluated not in silos, but by their cumulative lift in results.

Three puzzles, not one

The punchline is each type of behavior poses unique challenges. You can try to intercept consumers who mesh their related content, but be aware they may actually be using different content at the same time. You can also try to reach consumers as they stack different content on different devices, but to do so you’ll need to be more clever in how you coordinate your ad messages. And for consumers who shift across devices pursing related content, you’ll need expert measurement systems to understand this pathway and how to influence it.

After mobile: The looming future of screens everywhere

Screen Shot 2014-11-30 at 8.59.43 AM

Everyone is rushing to mobile and marketers want in. Facebook will clear $8 billion in mobile ad revenue this year, and Google will make $12 billion. Both have more than 1 billion users with access via mobile gadgets. Mobile, for a decade the Great Pumpkin of advertising, always unseen but about to arrive, seems to have finally emerged from Linus’s pumpkin patch.

But what if something bigger is looming behind today’s small-gadget lovefest?

That bigger thing may be digital screens, projecting images from any angle, wall or tabletop. At SXSW Interactive this spring, on a panel where Robert Scoble was still wearing his soon-to-be-discarded Google Glass, Gary Shapiro, chief executive of the Consumer Electronics Association, made a bold pronouncement: In a few years, he said, television screens will be as big as walls. Flat-panels will be everywhere. The corporate big-wigs will no longer be the woman or guy in the corner office with a window view, Shapiro said — instead, they will clamor for an office with a huge wall to install a massive digital screen.

Shapiro should know; his association is charged with researching consumer electronics trends and manufacturers’ product pipelines, so he skates to where the puck is going. First, the price of screen technology is falling. A 40-inch flat-panel TV cost $3,000 in 2003; the tag fell to $1,600 in 2007 and today, the same screen costs $330 at Best Buy. And second, screens are getting larger. This holiday season Vizio is selling an 80-inch TV for $2,499, the same cost of a panel half its size in 2005. Follow the trend line, toss in a bit of Moore’s law accelerating production, and if we can buy a digital screen that is 6-feet-8-inches diagonally wide today, by 2022 we’ll have screens that fill a living room wall.

But there is more here than just bigger TVs. The big story is the proliferation of screens and their corresponding input devices: the technology for making objects glow is spreading fast, and soon turning surfaces into screens may be as easy as painting an object. The image above shows glowing paper recently invented by Rohinni of Austin, Texas. Sony is testing watches made from flexible e-Ink paper. The gym I go to has a television image embedded behind the locker room mirror. And Disney’s research division has tested a Touche system that can turn any physical object — a tabletop or your sofa — into an input sensor that could control screens. Soon, fall asleep on your coach, and when your head hits the pillow your furniture could communicate to your home electronics to turn down the TV volume and dim the lights.

If the idea of paint that turns an object into a display screen seems science fiction, consider Chamtech Enterprises recently invented spray paint that turns surfaces into Wi-Fi antennas.

What this means is advertising communications in the near future will have far more screen options than the TV, PC or mobile gadgets most marketers are so obsessed with today. The myth of TV dying is just that. Mobile is rising fast, yes; Business Insider just published a fascinating report forecasting that mobile advertising dollars will make up more than half all digital marketing spending within four years, and noted that this year for the first time the number of minutes a typical consumer spends per day on mobile has finally eclipsed TV. (Note, television still captures more than 4 hours of viewing per person per day; the mobile devices are additive, not subtractive, in how people take in information as you “stack” your inputs between the big TV screen far away and gadgets nearby in your lap.)

Large screens offer a different experience than mobile, one more conducive to marketing. They tie into the third sphere of human psychological personal space, the distance  of 4 to 12 feet used for millennia as the story-telling field, the news your ancestors received from a campfire, a relaxing lean-back intake that we still enjoy in movie theaters or in front of basement TV sets. Personal space, as we’ve noted before, actually has three spheres of distance; intimate, up to 18 inches away; personal or working, 18 inches to 4 feet, the distance from our eyes to the tools in our hands; and social or news gathering, from 4 to 12 feet away. Mobile gadgets fit into the closest intimate field; laptops and computers and tablets the second working sphere; and large-panel TVs the third social sphere. For marketers, the larger screens in fields 2 and 3 provide much more room for exposition and storytelling, and consumers are more comfortable with unexpected ad intrusions in those social fields since they are not as close as our most intimate space. This core psychology is why ads don’t work well in mobile handsets but still do well in TV and computer browsers.

Take the long view, and mobile and its social halo could be a passing fad with a finite shelf life. Consumers have been mesmerized by such communication glitter before — telegrams, CB radio, long-distance telephone calls (remember them?), Second Life — only to see such manias fade. We already have glimmers that certain aspects of mobile may be declining, as tablet sales growth has stalled within only a few years of the iPad launch. Social networking, the communications bubble of the past five years, was recently dismissed by a Forrester report as a lousy form of marketing now being displaced by plain old banner advertising on Facebook and Twitter. Smartphones have turned into Star Trek communicators that do everything. But at some point, people may look up and see a new world of larger, proliferating screens.

When digital screen technology becomes so cheap that any object can be transformed into a glowing video image, the world of communications and advertising will unfold into a realm of infinite possibility. The challenge for marketers then may not be how to intercept consumers, but rather, how not to interrupt them too much.