Monthly Archives: June 2012

Apple patents a way to clone you (so your clone can lie)


Recently I wrote in Bloomberg Businessweek that someday you will have an artificially intelligent doppelganger, a clone of “you” whom you could send into work while you stay home to play golf. It would be relatively easy, simply combining three current technologies:

1. Voice recognition, like that in your car’s GPS system.

2. AI simulation, similar to Apple’s Siri.

3. Social media data sets, with all the content about your persona you upload into Facebook.

Obviously, a mechanical robot would be expensive, but with so much communication occurring over screens, it would be cheap to set up an image of your face, populate your fake mind, and like Siri, boot you up over Web communications or Skype calls to perform any service. 

On June 19, Apple was awarded a patent that would make such cloning real. Obscurely titled “Techniques to Pollute Electronic Profiling,” the patent spells out how to use a device to clone a person’s identity; set up the simulated intelligence in specific areas of interest; and — most interesting — process inputs to determine whether to act as you would, or respond by answering differently.

The Apple patent would create a cloned Siri-like avatar of you that could lie.

Why the lying? Apple notes “a significant concern with electronic commerce and the proliferation of electronic transactions is that of privacy.” People, organizations, businesses and government may be monitoring you, and this freaks some people out. Apple goes on to reference George Orwell and the Big Brother in his novel “1984,” in which this Big Brother sentient observation thing would watch people, and people would act in a way that was not true to themselves, but rather in the way they knew the government wanted them to act.

Apple suggests its AI-type avatar would block such monitoring by mimicking your behavior so observing entities will think the avatar is you … and then doing things unlike you to throw the observers off track. The patent states, “over time the eavesdropper will begin to associate divergent areas of interest for the clone as being the norm for a particular principal that the eavesdropper believes it is profiling or performing successful dataveillance on.”

Which raises lots of ethical questions. What will the world be like when you can send an electronic clone online to act like you, and do work for you (nice PowerPoint, avatar, I really like how you wove in the Wikipedia entry on “marketing”…)?

And once we’re in that world, what happens when your avatar begins to lie?

Corporate values, the dreamy blue-sky necessity

I recently took a trip to San Francisco and met with two companies, the first, a solar client, the second, the main SF Bloomberg office down on the wharfs. Both have open office environments. The solar company has low cubes for everyone including the CEO, and each conference room has floor-to-ceiling glass walls so any passers-by can see what’s happening inside. Bloomberg has an even more open environment, long tables of workstations with employees lined up before four screens each, a TV production area in the open center aisle surrounded by employees, and recording devices to capture what happens inside each conference room.

My first reaction coming from more traditional East Coast layouts was: Isn’t all this noisy. A deeper part of me wondered if such supposed transparency was driven by an Orwellian instinct to listen in on employees, avoid legal risks, etc. But I noticed something else — all the employees were more courteous than I’ve seen elsewhere. A spirit of shared professionalism was breathable in the halls. My Bloomberg friend and I chatted about tech column ideas and segued into a few colorful stories about sex, knowing that the conversation was going on tape. Who cared? The environment encouraged pure honesty.

The challenge with corporate values is like all intangibles, initially they are hard to tie to external customer demand or internal profits. Customers of that solar firm or readers of Bloomberg can’t see initially that the employees inside are acting more honest, more courteous, more respectful with each other, so “values” may not jack sales or readership. But that “value” bubbles up in productivity, idea generation, and service expansion in ways that help make such companies more competitive.

Corporate values in my mind are the fitness test of the production body, and when it comes time for the competitive race, companies that are more fit will perform better. Any company can optimize for itself in the short term by cutting corners and ethics, getting a bump in this quarter’s profits. Values recognize the deeper truth that the corporate, corporeal body really does not exist, and instead is a fiction filled with human beings whose ethical performance is tied to the product output. We’ve seen what happens to market ecosystems when values break down — housing bubbles, near-catastrophic financial collapses. The same lack of values can damage smaller corporate bodies.

It all comes down to motivating the real humans who make or serve things to perform, both for profit and for organizational strength. What is the value of “values”? What is the value of being surrounded by 100 honest, motivated, ideating, empowered, nimble, fast, trustworthy people who are pulling together in a common purpose? The second question answers the first.

Investments in fitness never pay off in the first day, but they do over the long haul.

(A response inspired by Gunther Sonnenfeld’s brilliant post on corporate ethics.) 

 Image: Sergio Tudela Romero

Microsoft plans ads that monitor your emotion

One of the fallacies of marketing is we “target” people as if they were static bull’s-eyes — you there, in the affluent demo, married with kids, PRIZM category mid-life crisis with SUV, intent to buy a red convertible — when really people contain multitudes. Today you’re an office worker. Tonight a mom. Later tonight, a Glee-watching high schooler. Much later, a sexy vamp. Whatever you’re into, you’re into different things, yet marketers fail to recognize your emotional and modality states.

Microsoft will change that. News percolated today that the Redmond giant has a patent for an ad system that would recognize your emotional state, and then match corresponding advertisements. The patent gives a hypothetical example: “For instance, OMG, Inc., is an advertiser that owns bowling alleys and lounges specializing in birthday parties…” OMG wants to serve ads that go “Bang!” with excitement, but kids or adults who are feeling sad don’t react well to such ads. So being smart, OMG sets up multiple versions of its advertising creative, and when the scanner/camera notices the digital viewer looks sad or stressed, either suppresses the ad (no whiz-bang party for you glumster) or could pitch an alternative (come relax in our chill lounge…).

Microsoft calls this “emotional targeting.” You may ask, how in the world could a company track our emotions? Well, your laptop has a camera pointed at your head; Microsoft owns Kinect, which uses body-monitor scanning to allow cool video games, and Skype, where you share videos of your face; in fact, every company working on a smart TV is embedding video cameras aimed directly at you.

In George Orwell’s 1984 the TV sets watch people. In one scene the protagonist moved out of the room, around a corner, to open a book or journal to try to have privacy, but the monitors caught him. Are we afraid of what could happen if the devices we love to watch start watching us back? Or is this a natural extension of marketers trying to personalize offers meaningful to us by moving beyond data trails to direct observation?

If you’re curious, here’s a list of the ways Microsoft says it might track you:



The client devices … include, without limitation, personal digital assistants, smart phones, laptops, personal computers, gaming devices, or any other suitable client computing device. In some embodiments, the client devices include image capture and voice capture devices. The image capture devices include cameras, video cameras, etc. The voice capture devices include microphones, recorders, etc. The client devices … include a user and system information storage to store user and system information on the client device. The user information may include search histories, cookies, user identifiers, online activities, assigned emotional states, and passwords. The system information may include Internet protocol addresses, cached Webpages, and system utilization. 

Apps as disposable media

Famed 1990s’ bubble analyst Mary Meeker is out with her annual digital media forecast, and one of her jaw-dropping findings is that Apple users now download a collective 46 million apps each day. At first, reading this, you go “yay, apps!” And then you pause. “Crap. That’s a lot of apps. How can people use so many of those software-ish things?”

And that is the problem. Apps are no longer software; they have become commoditized, fly-by-night media. Apple has, by our estimates, 300 million current iTunes accounts with registered credit cards. If you divide 300 million users into the 16.8 billion apps downloaded each year, each Apple user grabs 56 apps annually — or about one per week per consumer.

No one uses 56 apps. Which means apps are disposable. Instead of a software portal (each app-maker’s dream) or new platform (you want your app to become the next Foursquare yes!), apps are now just a media slot — easily seen, quickly forgotten, like a TV commercial or banner ad flashing by in the night.

This doesn’t mean you shouldn’t build an app. Instead, it means you need to launch 100 of them. Apps get noticed — users love to download them after all — but their lifecycle is short. So go build an app. Then forget it, because next week you’ll need to build another.

Brilliant ideas are commodities (Part 1)


What I find interesting is that within ad agencies, 99% of the conversation about what they do is focused on “the creative idea,” and yet outside of ad agencies, 99% of what people talk about in communications is the change in media channels. Go on Twitter and try to find 20 tweets about an advertising creative idea that rocks in the next 5 minutes; good luck (Super Bowl week doesn’t count). Go on Twitter next and try to find 20 tweets about what Apple is doing with its upcoming hardware gizmo, and you’ll go “check” in a few seconds. People are obsessed with media channels and tools; agencies keep talking about ideas. Therein is a disconnect. The intersection of media channel and creative idea is rarely explained, although a few smart shops such as BBH Labs have discussed ending the top-down creative-to-media funnel and building a more integrated planning whole.
Or put another way, good creative is now a commodity.
The reason I suggest this is in a world where people watch 5 hours of TV a day seeing 166 :30 second spots, or spend 3 hours on the Internet a day being exposed to thousands of banner ad and video pre-roll impressions, 99% of creative ideas are ignored. Completely. Creative is really now a pass-fail grade — you get noticed, or more likely, you don’t, and even if your idea is in the 10% of brilliant executions, you’re still competing with 16 other top-of-the-heap TV spots and a few hundred other banner ads.
Go ahead, build a brilliant idea. You’ll be one of 50 or so I’m exposed to tomorrow.
This is not a negative; advertising is as always a game of what we catch, not what we spill. You could argue as well that good ideas and creative are more important than ever before in a world agog in communications overload. But this commoditization of idea brilliance is a real problem, and you can see it yourself if you measure the downcycle timeline of any big “idea” that goes viral. Most newfound memes spike but for a few days and then disappear. See ya, Skittles homepage.
What I’d love to see, if I were a student in Edward Boches’ class on creative, is how the construct of media channel and the pinnacle of brilliant creative inform each other, instead of the “idea” itself being something separate. This combination is the only path I see to truly breaking out and building something with sustained power and resonance.
Reposted from my comment here


Image: Nixter

When the air itself becomes the gadget

One irony of our virtual-networked age is consumers are still gaga about gadgets. The Internet and apps may give us a million different ways to view weather forecasts on a screen, but as soon as Apple launches a thinner MacBook Pro Air with a black bezel, we’ll run to the mall.

The challenge of course is computer product designs are converging into flat panes, and eventually panes can only go so far. When screens and smartphones achieve the apex of glass, product differentiation will be difficult. Which is why devices soon will move out of solid shapes.

Two examples are laser keyboards and miniature projectors. The Cube Laser Virtual Keyboard is a $180 gizmo that beams glowing keys onto any flat surface, and somehow tracks the position of your fingers as you “click” on the flat QWERTY layout. You pair the device with an iPad and suddenly can type away like mad. (Flatscreen tablets suck at typing, yes.) It doesn’t take a rocket scientist to guess that within two years Apple and Samsung will add such laser-keyboard inputs into their tablets and phones. And for output, miniature projectors do exactly what they sound like — beam images from your phone and tablet onto the wall, so you can regale dinner companions with cat videos or hold an impromptu PowerPoint presentation with that executive you meet in the bathroom stall.

The third and most promising way devices will leave their hardware shells behind is virtual reality projections. Google announced this week it is expanding its Google Maps 3-D modeling (which renders photorealistic images of major metro buildings, streets, water, and flora from aerial imagery) to mobile phones. Now your handset can unveil a virtual earth tied to your location. If Google has figured out how to compress this powerful software into small handsets, the next step will be putting it inside your glasses, and soon you can overlay any fiction on the world you wish. Some clever hackers twisted the Google Project Glass teaser video to show how you could overlay the “Battlefield 5” game onto your neighborhood walk, if only you wore the right pair of virtual-reality spectacles.

Soon, keyboard inputs, video projections, and virtual reality will dance in the air around our fingers and eyeballs. The hunger to buy the next Apple product will fade, because slightly recast aluminum shells will become commoditized and a glass tab that transforms into a high-def screen is just another piece of glass. Apple, Google/Motorola, Samsung, Dell, HP and other gadget manufacturers will need to spend more time thinking through virtual interfaces than concrete shells. Play it forward and you’ll see plenty of opportunity for garage startups to break into this new anti-product world. When the air itself becomes the gadget, the definition of product design will change.

We tested ‘engagement’ on Facebook. It wasn’t pretty.

This is a story about how I flirted with Zappos, yet Zappos did not return my call.

Social media is supposed to allow brands to build “engagement,” a quasi-relationship that relies on frequency of interactions. Ian Schafer, chief of the sharp social agency Deep Focus, wrote an Ad Age column recently saying frequency is key to setting Facebook apart from just another advertising medium. Ian commented, “Facebook enables frequent interactions with consumers over time … real consumer connections become the new impressions.”

Why is this important? If you want to influence someone, you have to try repeatedly. A classic precept of advertising is that to influence a consumer, you must touch them with a “frequency” of three communications per week to get the most response. The big soap-makers (P&G, etc.) in the 1970s and 1980s conducted numerous studies that found 3x contacts per week is just the Goldilocks amount of outreach to get the most people possible to respond to you. This is rather intuitive; your spouse didn’t marry you on the first date, so one “frequency” wouldn’t have worked in your courtship, yet if you hounded her every hour of each day initially, she would have reported you for stalking.

So I decided to test how well brands use frequency inside Facebook. I “Liked” 12 brands on Facebook including Apple, Bank of America, Ford, Pepsi, Starbucks, Zappos and others, as well as the U.S. presidential candidates Barack Obama and Mitt Romney, and then carefully checked into Facebook 31 times over the next 7 days to see how frequently they reached out to me organically in my stream. Liking a brand on Facebook, of course, gives that entity permission to pop up in your Facebook feed like an old girlfriend. Each time I checked into Facebook, I scrolled back for several hours of updates, fervently looking for engagement from my newfound brand friends.

And what I found is … most brands fail at using social media to build frequent interactions. Only Mitt Romney and Barack Obama did well, reaching out to me 5 and 4 times respectively. But the commercial brands averaged a frequency of only 0.6 interactions over 7 days. Less than 1 interaction — far below the threshold of 3 touchpoints required to influence a consumer! Several brands including Apple, Bank of America, Starbucks, Trek and Zappos didn’t even show up once in my Facebook feed.

Personalization? What personalization?

The other big problem I noticed with Facebook engagement is when brands did send a missive into my stream, it was one-size-fits-all. Pepsi gave me an extended version of its current TV commercial (the irony). University of Phoenix popped up offering me a teaching certificate, while my bio clearly says I work in advertising and marketing. When a brand did try to “engage,” it felt about as on point as a blaring AM radio commercial.

This isn’t really a ding on the Facebook platform as much as worrisome that leading brands such as Pepsi and Zappos can’t build coherent relationships with new fans. Big companies seem more intent on using social platforms as a broadcast medium than a two-way communications stream. But this could also indicate Facebook needs to provide better tools to enable one-to-one personalization, or quasi customization, in brand outreach. If I like Trek, maybe my updates talk about mountain biking more than road cycling, so Trek should be able to tailor its missives to my interests. Spraying all your brand fans with the same message feels a little mass media to me, and while that can work, is it what we really expect from social media?

So here’s an idea. Instead of getting me to Like your brand, how about having your brand Like me?

Ben Kunz is vice president of strategic planning at Mediassociates, an advertising media planning and buying agency, and co-founder of its digital trading desk eEffective.


Image: Pulpolux

On the fallacies of attribution (making customers love you)

In a perfect world of cause and effect, every advertising action creates an immediate reaction. Since Force (result) equals Mass (amount of input) times Acceleration (rate of change), marketers who know both inputs M*A should be able to predict F — and thus force the result. Decades ago, such analysis didn’t matter much in advertising, back when magazine copy and radio jingles were king, but since the rise of digital media, where every nuance can be measured to a microscopic level, marketers now are expected to map everything to cause and effect.
Did you spend $10,000 on sky-writing “I love you, customer” for 1,000 sales? Congrats. That’s a $10 cost per sale. She loves you. She loves you not.
However, this is wrong, at least in many areas of advertising. Consumers are not just trigger-response lovebirds, and not all marketing can be evaluated on a pure kiss-or-no-kiss basis.
The challenge for marketers is that consumers are not primed to respond like a teenage boy on prom night at the first stimulating impulse. To paraphrase the great Don Peppers, consumer desire is not a binary switch; instead, it is a volume dial.
The better metaphor for marketing response is interplanetary gravity. Customers circle your product or service at many distances, from the remote awareness out at Pluto’s orbit to the closer asteroid belt of consideration to Mercury’s rapid inner circle of intent to the inbound plunge of a comet making a purchase. The sun’s first tug on an object far outside our solar system creates a complex, slow journey inward, and if that object doesn’t immediately fall into our star, we should not say gravity has failed.
Our media agency is asked to solve this challenge all the time — does advertising work, will it work, if we spend $1,000,000 on advertising medium X what will the immediate response Y be? Many campaigns can be predicted at this level, and tracking each channel’s performance vs. its peers can lift results by 50% or more. Yes, advertising forecasting and measurement is critically important, and we do it with top-of-funnel direct-response tags to bottom-of-the-cycle CRM system integration.
But … a myopic focus solely on direct response ignores the need for consumer education, product illustration, message exposition, and relationship nurturing. Relationships, after all, have beginnings and middles before consummation, and the courtship between any two entities is typically a delicate affair.
Did you buy a car the first time you thought of an upgrade? No.
Did you grab the first smartphone you saw immediately after learning that phones could act as mini computers? No.
Have you ever bought anything several weeks after seeing the product ads, only after the gnawing hunger of desire worked its way through your consumption system? Of course, yes.
Pop quiz: Your vacuum cleaner breaks next Friday night, one day before you’re hosting a huge party. You rush out Saturday morning to buy a new vacuum. Are you interested in one of the newfangled bagless brands you’ve seen advertised months ago? Check. But any marketer measuring those print or TV Dyson ads from those many months ago likely was disappointed in the immediate sales results — if responses within the next week were expected then.

Linear modeling works in advertising. It is also only one type of modeling.
Linear modeling means drawing a line between two things, and it is a superbly simply math formula that fails on most levels. Let’s go back to love. If your spouse measured the ROI on you the night she went home after meeting you for the first time — perhaps with a smoldering glance of eyes and mild chemistry but you didn’t do a damned thing, no engagement ring, really? — she’d check off “zero” on a spreadsheet and cancel all future communications. 
A better solution is to look beyond stimulus A and end point B, and map a series of intermediary milestones that your newfound love must reach — and be re-stimulated in. A simple scenario:
  1. If a customer lead is generated at $100 and you have a 10% conversion rate, you have a $1,000 cost per sale. 
  1. If you spent another $20 on two additional marketing touches in an interim step on the same customer lead — say, an outbound phone call, and a nice direct mail piece with a product DVD — your cost per lead would rise to $120. But if those two additional touches boosted conversion rates from 10% to 15%, your cost per sale falls ($120 cost per lead / 15%) from $1,000 to $800. 
With a little more effort, you’ve achieved a 20% improvement in advertising performance. 
Consumers have modalities. Their urges are held in restraint. They learn about you long before they want you or need you.
You aren’t selling the customer. You are building a relationship with a customer, and even if that connection will be fleeting — climaxing in only one sale — in the early stages, love takes time.

Ben Kunz is vice president of strategic planning at Mediassociates, an advertising media planning and buying agency, and co-founder of its digital trading desk eEffective.