Google and NYT knock out software by watching you

Last fall we came across Google's image labeler, a little game that invites you to race a stranger (somewhere out there) in tagging photos with titles that make sense. We got served Drew Barrymore and typed "knockout."

Turns out Google and other companies are using your personal down time to improve how computers recognize photos, video and scanned text. Humans are better than computers at image recognition; but if millions of humans say an image is X, the computer begins to get it too. In the cleverest move, the twisted-word Captcha codes you type to gain access to Twitter or Facebook are being monitored by The New York Times to improve computer recognition of printed words ... in essence, using 10 seconds of your brain to refine software that will scan back issues of NYT from 1851 to 1980.


Both ideas, the image play and "ReCaptcha," are brainchilds of Luis von Ahn -- a Carnegie Mellon guru who created the fuzzy password tests for Yahoo in 2000, and is expanding to use downtime to solve problems of artificial intelligence. The average U.S. consumer spends 1.1 hours a day on electronic games and 1.7 hours using email, all input-heavy interactions that could conceivably be leveraged for broader computing tasks.

Just think of what he'll do with the 17 minutes you spend in the bathroom.

Tip from Brad Ward.

If your readers fell in a forest, would you hear it?

WSJ.com proves that affluent readers continue to move online. In March unique visitors at the site were up 175% to 15 million, despite the fact that WSJ.com continues to charge for much of its content.

Meanwhile, traditional paper-based news magazines are tipping over. Newsweek's circulation is down 16%; Time has plummeted 19%; and U.S. News & World Report announced last week it would cut its rate base a whooping 500,000 to 1.5 million -- and print only 36 issues a year instead of the former 46.

These numbers mean that anyone advertising in traditional print may be seeing declining inquiries and rising customer acquisition costs. If you have been reluctant to explore web display advertising, it's time to start that engine. (Photo by Cicada.)

Behavioral targeting: What you want, and what you don't

Behavioral targeting is the advertising approach of monitoring a computer user's behavior across dozens of web sites, and then serving up targeted ads in response. Read about finance, and you may soon see banner ads for insurance.

It works brilliantly, creating response rates up to five times higher than regular web site banner ads -- but now we find consumers aren't too happy about it. A new study by TRUSTe found that while only 40% of U.S. adults were familiar with the term "behavioral targeting," most knew what it was -- and didn't like it. 91% said they would do anything to ensure better privacy online, and 42% said they would gladly opt out of all tracking even if it meant they never saw a relevant ad.

Behavioral targeting has been getting heat; the Center for Digital Democracy and the US Public Interest Research Group complained in November that the practice was "invasive" and caused "new threats" to children.

Which poses a conundrum. Consumers say they don't like it, but consumers respond much better to targeted ads. It's a bit like knowing chocolate is bad for us, but taking a bite anyway. There's a difference between what people say and what they do.

Pull your web site out of the Briar Patch

This week we met with several executives to discuss how to acquire new customers online -- the usually mix of offline media, Google, and ad networks pointing to the mission-critical web page. We were met with enthusiasm, plus the common barrier of, well, we already have a web site and the development team is working on it ... meaning, nothing may happen for months. Their web site is caught in a Briar Patch.

The Briar Patch, you recall, is the maze of thorny bushes to which Br'er Rabbit escapes in classic Uncle Remus stories. The tale comes from the American South, but actually has its roots deep in Central Africa, from folklore trickster stories in which the rabbit represents someone faced with adversity who uses his wits to beat a snarly puzzle. And if that doesn't sound familiar, well, you've never sat in a web design committee.


The fastest way out of a web tangle, we suggest, is to do what leading companies do -- create a simple microsite, test advertising to drive traffic there, and see what you get. Great examples abound among automotive and oil companies, which are falling over themselves to reposition their cars or fuel as environmentally friendly. The benefits of a nimble microsite:

+ You don't have to reposition the entire company to meet a specific need ("I'm concerned about the environment")
+ You don't have to change your entire site to go live ("Hey, look, a clean simple site that links back to the main one")
+ You attract a self-selecting audience (only people searching for this topic will find the microsite)

Toyota, Chevy and Shell all have "green" microsites. Never mind that some of the claims are simple greenwashing; Chevy, for example, brags that its hybrid Tahoe SUV was named Green Car of the Year for getting 22 MPG highway; you can drop a Duramax diesel stock engine into a Hummer and get 22 MPG also.


Doesn't matter. These sites attract only auto enthusiasts trying to appease guilt about their carbon-emitting ways. The ads driving traffic appear in online sites targeting affluent, educated, liberal readers (David Pogue's column in NYT) or in glossy print targeting the same (Harper's). Only those who really care will find out that Volkswagen offered to offset your carbon footprint.

It's a good case study in tailoring the message to a subset of your audience. The web is one key to converting specific consumer interests into sales action. You don't have only one brochure. Why in the world are you hinging your online hopes on only one web site?

Beware of GVI: Google Visual Inventory problem

The more we think about the Google Visual Inventory problem, the more we realize it threatens all internet advertisers. It's simple. As consumers shift to mobile devices, interface screens get smaller -- leaving less visible room for ads. Google, and other advertisers, run out of visual inventory.

Ad inventory is the bread-and-butter of making money online; it is why it takes you three clicks, through three screens, to find local weather information at weather.com. More visible real estate = more ad space to sell = more revenue, so sites like weather.com have an incentive to push each user through large, multiple screens of data. Today Google crams nine or more ads from search results on a PC screen. Alas, only one ad will fit on most mobile phones. As millions of consumers adopt mobile internet in the next few years, Google and advertisers will have a hard time finding shelf space.

(Now, if Google were really worried, they'd probably want to start designing a new cell phone interface ... oh, never mind.)

The GVI challenge is one of several cracks under advertisers' feet. The second serious challenge is that online users migrating to social media are no longer paying attention to ads, since they are focused on engaging socially with friends. Sure, Slide may reach 140+ million users with widgets each month, but if you write us with the names of three ads you recall seeing while using SuperPoke or FunWall, we'll send you a prize. And now Danah Boyd notes a third scary trend for online marketers: not only are social media users ignoring ads, users are beginning to lock unwanted messages out. In a recent speech Danah

... basically told this room full of marketers desperate to get on teens' friends lists, that those teens think that's creepy and invasive. Why? Because it's THEIR Space -- even if it's public (which most teen profiles aren't anymore). She had a great analogy of teens telling their parents, "It's MY ROOM." And the parents telling teens "It's MY HOUSE." Just as teens put "keep out" signs on their bedroom doors, teens have created "structural walls" to keep everyone but their friends out on social networking sites. (Tx Anastasia Goodstein, for the report.)

Ouch. Less ad space. Less-attentive audience. Consumers doing something other than listening, and consumers more than willing to block out your peripheral pitch. GVI is more than a physical screen problem; it's a metaphor for the new mindset of the MySpace generation who are crowding out ad messages with their own portals to private communications. Egad, marketers -- suddenly logos on cotton T-shirts are looking fine.

Add 'microsite' to your 2008 to-do list

One of our goals in 2008 is to encourage clients to rethink their web landing pages. As media planners, we focus on the front end of advertising -- reaching specific demographic groups most likely to want your product, for example, men age 35-44 who are weekend warriors likely to blow out a knee and need orthopedics, or the women in their lives who guide the majority of healthcare decisions.

We don't build web pages and so have no vested interest in saying this: But let's face it. Your web site probably needs work.

The challenge with advertising, especially advertising online, is that you can lead a prospective patient or customer to your web site -- but what happens when they get there? Many web sites are ill-structured to convert a visitor to a qualified "lead." And web sites are not always easy to modify. Smaller businesses don't revamp web sites often, and large bureaucracies often find changes mired in Steering Committee or IT meetings.

The quickest solution is to build a microsite, such as the bariatrics site above. This concept is simple -- build a small, nimble subset of your brand, focus it around a specific customer or patient need, and launch quickly. Microsites have several benefits:

+ You can typically complete the entire project within 6-8 weeks
+ You avoid nasty internal debates over how to improve the current vast main site
+ It extends your reach on the internet, creating more "points of entry" for customers
+ You can rapidly add a visible lead form -- to collect the visitor's name, email, phone and ZIP -- to build a prospect database
+ You can show your boss, hey, look, we actually did something this year (Admit it. That would feel good.)

Some marketing executives shy away from this concept, certain that their master web site is enough. This isn't the case. If you pull a report of the first pages visitors hit within your web site, you'll find the majority are landing deep inside -- because most web users start at Google, and Google's search engine throws them past your home page into the innards.

A microsite replicates what people want when using the internet -- finding information rapidly about what they are searching for. Worth considering in 2008. Plus, it makes internet and media planners like us much happier when the leads we deliver to your web site find a simple way to give you information, so you can turn them into customers.

Why widgets are a bit like Frisbees

Tom Giles, technology editor over at BusinessWeek.com, has put together an intriguing debate on whether widgets are the future of online advertising. Widgets, as you know by now, are little online applications where web users can play around with graphics and data, and potentially pass them on to thousands of their friends. If a web banner ad and video game had offspring, it would look like a widget. Marketers are pretty excited about this stuff because all those young users inside MySpace and Facebook have been ignoring traditional banner ads, and widgets offer another way to grab their attention.

Some widgets, like the iLike music app, take off -- but most do not. So, to play maverick, we gave BusinessWeek a column saying widgets are vastly overrated. Sure, it's nice that your ad can now wiggle on a computer screen, but we caution, marketers don't buy advertising -- they buy customers. And the only way to acquire a customer is to reach someone when they are paying attention.

The trouble with widgets is they are a bit like a Frisbee being tossed around by young women at a park. You could put a corporate message on top of the plastic disk, but will those users really remember it? And if those women tossed you the Frisbee, would you pay attention to the Frisbee's logo -- or to the new friends you just made in the park? Consumers have different modes, and if you reach them when they are not receptive to a marketing message, you've just wasted an impression.

Widgets can work, but only if they meet four criteria. Check out our point of view in BusinessWeek.

Google slippage, or how to predict a decline

So analysts bashed Google stock today after seeing that pay-per-click ad revenue is slipping. We've noted that Google Trends shows global searches for common phrases are sliding as well, with demand down about 50% in the past four years. Seems Google is also playing with its formulas to wring more money out of click budgets -- a potential sign of desperation. Google may be in decline.

Which brings up a point: Why can't anyone see changes like this coming? Take any infatuation -- AOL, Google, MySpace, Facebook, widgets, real estate, the 17th century tulip craze -- and people jump in as if the ride will last forever. But the reality is almost everything in life follows a bell curve path toward the future. By the time you hear something is hot, chances are that same something is about to peak.

Look at your own life. If you work in marketing but aren't CMO or agency president yet, you're probably in your 30s or 40s, have rising income, good job prospects ... but chances are, eventually your earnings power and title and circle of colleagues will peak, and then diminish.

In marketing, clients and agencies are always hungry for an edge, so they tend to disregard the future downward slope and leap into every bubble. We recently had the opportunity to talk with BusinessWeek.com about widgets, and the reporter mentioned that everyone who produces these things proclaims them to be the future of marketing. Uh-huh.

Don't get depressed. Just, when planning your future, remember that what goes up must come down.

It's not just Google. It's a complex mousetrap.

Mario Sgambelluri throws cold water on the whole Google-has-the-best-results thing by pointing out the way most people measure internet campaigns disregards everything we know about marketing.

The logic trap marketers fall into is thinking the last ad people saw triggered the response. So if a text ad on Google creates a click at 25 cents, that's good, right? Sgambelluri points out that people take actions based on a culmination of impressions. The concept of integrated marketing has been around for decades, yet we seem to forget this when evaluating online ad components. A Google ad helps close the deal, but prior impressions in offline and online media helped get consumers to the point of decision.

This is important, because as marketers allocate funding in online campaigns they must consider the upstream online communications -- banner ads, blogs, social media, videos, microsites -- that lead consumers to the search engine trigger point. Atlas Institute source presentation here.

Zap! Now ads are vaporized on Safari, too.

So 2.5+ million people around the world have downloaded Adblock Plus, a web browser add-on that blocks banner ads, and adoption is climbing at about 300,000 to 400,000 users per month. Last year it was only available on Firefox, but it got enough buzz for PC World to call it one of the top 100 products of the year.

Now, an adblock is available for Safari web browsers, too. Screen shots below show a French web page, before and after banner ads are vaporized. As consumers begin screening ads, as web modality shifts from search to social, and as the screen inventory for ads gets smaller (see: your cell phone), internet marketers may face rough seas ahead. Expect online ad channels to fight back -- like the move by Google to begin testing video ads in its search results pages. If the ads are sexy enough, who wants to block them?

Big, bad news for web ads: 6% of users drive 50% of clicks

A new study by Starcom, Tacoda and comScore may shake internet advertisers to their core. Seems only 6% of online users create 50% of all clicks on web banner ads. Worse, this "heavy-clicker" group is not a sweet demographic target:

- Heavy-clickers have low household incomes below $40,000 a year
- Heavy-clickers spend more time online, but this is not equal to higher online spending
- And, in a death knell for "but wait we're just branding" campaigns, the study found there is no correlation between high clicks on banners and increased brand affinity.

Put it together, and you get a bleak picture. Your click-through rate may be above average, but you may be getting clicks from just a low-income, brand-oblivious fraction of your total online target audience. This doesn't mean that banner campaigns don't work, but it does mean that click-through rates alone won't tell you results. You'll have to track conversion rates, costs per lead, costs per acquisition, and even new customer profitability and lifetime value ... the deep metrics that explain the real web campaign results.

Perhaps high-income, highly educated web users have better things to do on the web -- like use the web.

The Root fulfills dream of national black newspaper -- only online

If you're interested in the politics of race, guruship by Malcolm Gladwell, African-American genealogy, and Obama vs. the Clintons, then you'll like The Root, a new web-only publication by Henry Louis Gates Jr. and Washingtonpost.Newsweek Interactive.

The Root launched yesterday with star power, an innovative business model, and close promotion with Slate.com. Co-founder Gates notes that "since 1827, black journalists have dreamed of creating a national black newspaper." The web magazine offers political and cultural commentary, plus a unique tie-in to AfricanDNA.com, a genealogical charting and mapping effort. And in a twist sure to get buzz, the service allows readers to send in their own DNA samples for analysis of their country origins.

The strong web-only stance is going to get noticed. The closest competitor we see reaching out to affluent black consumers with similar news and analysis is EbonyJet.com. Advertisers, take your marks.

The rare brilliance of Darryl Ohrt

We say rare, because we know this guy, and he's not always so articulate. But damn. Don't miss Darryl Ohrt of Plaid being interviewed by Laura Newman on the future of the internet, advertising, and how agencies should work with clients. He must have had a Red Bull, because this is genius.

Our favorite bit is Darryl talking on how business strategy must evolve:

"The internet has changed our relationship with our customers dramatically. Some have said 'Google is God.' Meaning that Google is all-knowing, and finding everything that you do. You should expect that everything about you and your brand will eventually be accessible via the internet. Run your business accordingly. I believe it comes down to three simple words:

1. Honesty. Be proud of who you are. Accept your shortcomings. Don't try and be someone or something that you're not. If you're a small company, be proud of that. People will love you for who and what you are.

2. Integrity. When you make a mistake, admit it. Stand behind your products or service offerings. Do it well, and your customers will become evangelists.

3. Sharing. Blogging, Twittering, Facebook - all excellent tools to share with your audience. But it's more than just using the tools of the moment. What are you sharing? Is it of value to your audience?

You'd think that these three things have been a part of the business community for ages - but it's not so. The internet has put every pixel of information at our fingertips. Which means that the brands who are trying to hide something, or twist a story will always be found out. The age of 'spin' is over."

Microsoft's next ad targeting: Online, offline, everywhere

A new Microsoft patent application shows the future of advertising, and it has U.S. privacy advocates worried. The system would combine data from cell-phone towers to pinpoint a consumer's location (by triangulating on your cell phone); data from credit cards about transaction history; and any media the consumer is now using, such as watching TV, listening to the radio, or potentially strolling past a digital billboard.

The application says:

"If the offline behavior indicates the user was watching a college football game ... if the user goes online during or just after such activity, then an inference could be made that the user is interested in seeing more information about the game as well as being receptive to advertisements selling college-team memorabilia."

Soon advertisers may be able to move beyond ad networks to purchasing people networks. "Hey, we want to target James Smith. Let's track his every location, and serve him ads at every media outlet." Question is: if personalization becomes omnipresent and omnipotent, won't consumers get sick of it?

Banner ads for the little guy

Now AdReady offers anyone with a credit card the ability to customize a web banner ad template, run it over an ad network of hundreds of web sites, target specific demos, and only pay per click. The ad templates don't look so hot, but then, neither do most web site layouts. Plumbers, hairdressers and lawyers, once you have Google down, this is the next step.

Gotta go. Our umbrella is calling.

Now umbrellas are connected to the web for realtime weather info. LG is bringing back the internet refrigerator, with recipes, 5-day forecasts, and reminders for your wedding anniversary. The nimbus of information now in your browser window is breaking loose and surrounding every ho-hum product with a hip halo.

Question is, when every toothbrush, hairdryer, and auto windshield display has a direct GPS feed into the internet, how will advertisers intercept viewers focused on mundane tasks? Will we want banner ads when we reach in for the carton of OJ?

Answer is, no. As consumers spend more time getting information from direct feeds into each product -- and not web browser windows -- the hours spent on the traditional internet may begin to decline, just as broadcast TV, radio and print are in free-fall now. In 5 or 10 years, the already fragmented audience for advertising may move away from the current online browser bubble and into a world of utilitarian product data. What Twitter did for email, umbrellas may do for banner advertising.

Back when banner ads had no clicks

Just when we're feeling blue about Apple not releasing a 3G iPhone, we catch images like the one above of a sick migrant child worker in Yakima Valley, Wash., in 1939. This weekend Andy points us to Shorpy, a brilliant compilation of 100 years of American photography. The poverty and austerity of the earlier generations provide striking perspective on our silly gadgets today.

Many images show a country pushed to the roads. The dust bowl and depression of the 1930s set America in motion. People lived in tents; entire families climbed into rickety railcars; autos out-raced dust storms.


And the era created outdoor advertising -- a plethora of postings for Chesterfield, Coca-Cola, Burma-Shave, trying to catch people as they approached country stores. In some ways, it was a golden era of outdoor, before the interstate highway system of the 1950s took people away from villages and made small signs impractical.


Strange thing is, all the little signs of the 1930s remind us of today's banner ads trying to catch eyes on the internet. New highway. New store fronts. Same tactics.

Dear Jakob Nielsen, why are we hiding good web sites?

If you're not aggressively using search engine marketing (paid ads on Google, Yahoo, and other engines) to help customers find you online, here are five reasons to consider it for the new year. (Tx to usability guru Jakob Nielsen for inspiration, although he prefers to focus on web design.)

1. There are more than a billion users on the internet. Nielsen notes that if your web site attracts fewer than 10 million users, you're missing 99% of your potential audience.
2. Most users start at search engines. For example, 66% of U.S. consumers searching for healthcare information online begin at a search engine, not a specific site.
3. Nielsen’s usability studies found that 60% of initial page views were deep within a web site, not the home page ... meaning users landed from search engines or direct links from other sites to specific information.
4. Most sites attract one-time users. A site has only a 12% probability of being revisited. So you're better off casting a net on search engines for new visitors rather than trying to make current ones stick.
5. Google made $4.2 billion in revenue in the third quarter of 2007. No one makes that kind of cash unless their service works.

Start yourself here. Or get professional help here.

2008: Netflix moves away from red envelopes, broadcasters worry

Another trend to watch: the slow death of broadcast television. NYT reports Netflix may eventually ditch its red mailing envelopes (used to mail movie rentals to U.S. homes) to broadcast video directly to high-end TVs over the internet.

LG could integrate the Netflix service into a future version of its dual-mode HD DVD/Blu-ray DVD player, which now sells for $799, and a new line of high-definition TVs with wireless connections to the Internet, among other products.

So Netflix will soon be another broadcast channel, in which you rent a movie by clicking on a screen. Extrapolate the trend line and you'll see trouble on the horizon for broadcast. As consumers gain more control by "pulling" video content whenever they want, the old broadcast advertising inventory will have less and less impact.

We've already seen this trend in print, as the movement of readers to free content online has created a vicious cycle of diminished ad response, reduced demand for advertising, fewer advertising dollars, and eventually shrunken or dying newspapers.

What is the solution? As media fragments to the nth degree, advertisers will need to move to networks of users and not ABC or CBS. The ability to target will be much higher, but the complexity of the buy will be enormous as advertising options tied to video morph over the next decade. We don't expect broadcast to die entirely, since consumers often just want to relax and have the cold blue light of cable wash over them. But a large part of that passive audience is going away soon.

2008: The year internet ads will outpace radio

Deep on page 126, The Economist offers an illumination:

The internet will account for 8% of global advertising spending in 2008, says London-based ZenithOptimedia. For the first time, the web will outperform radio, which will account for 7.9% of global advertising outlays. The internet is growing six times faster than traditional media, and by 2009 will account for more than 10% of ad spending in 11 countries, including Norway, Sweden and Britain.

If you work in marketing and are not yet testing online ad formats, please, get started. The train is leaving the station and your competitors are on it.