About Frédéric Filloux

http://

Posts by Frédéric Filloux:

Off The eBook Shelf

Readers are voting with their wallets: The eBook is winning. In the US, eBooks sales are now topping hardcovers for the first time (story in TechCrunch). Not everywhere of course. According to the Bowker Global eBook Research, the global market for eBooks is driven — in that order — by India, Australia, the UK and the United States. The laggards are Japan and (no surprise) France. The chart below shows the percentage of internet population reporting the purchase of a digital book over the last six months prior to the survey.

Interestingly, for most population samples, the level of purchases is not correlated with awareness. France enjoys the highest level of awareness but its internet population buys five times less eBooks than India’s. Once an Indian internet user finds an attractive digital book offer, he/she will most likely jump on it. This could lead to the following: in emerging countries, the cellular phone has become the main communication tool, leapfrogging the deployment of land lines; similarly, we could see eBooks bypassing print in countries like India where a large segment of the population is getting both literate and connected at a fast pace. (Actually, Bowker also reports that over 50% of respondents in India and Brazil are likely to buy an eBook in the next six months, ten times more than in France.)

If the rise of the eBook happily provides access to knowledge in emerging countries, the picture is more contrasted in countries with a long history and high penetration of printed books.

For instance, let’s have look at the ISBN registrations data for the United States. The chart below, drawn again from Bowkers (full PDF table here) shows a steady progression:

Between 2002 and 2011, in the US market, ISBN registration grew by 61% and reached 347,178 new titles. (A technical note: I’m only taking into account books that fall in an identified category, such as Arts, Biography, Business, etc. I’m excluding the huge segment labeled as nontraditional, which includes reprints, public domain, and titles printed on demand; this segment grew by over 3500% to 1.2 million registrations, which would distort the picture.)

We clearly see the impact of mainstream e-readers such as the Kindle and the iPad. Without any doubt, they contributed to the growth of registrations. (Unfortunately, ISBN counts does not provide a breakdown between print and digital.) Over the last nine years, some book publishing segments fared better than others. See the chart below:

Fiction is doing twice better than all other categories together. The Digital Book is the medium of choice for fiction: a) eBooks are set to be cheaper that print and price elasticity is now a proven fact, the cheaper a book is, the more likely a reader is to try it; b) e-commerce breeds impulse buying (cf the famous “One-Click® feature); c) readers can test the product more efficiently than in the printed world as Amazon and the iBooks Store make larges sample available for free. No surprise, then, to see the Fiction category holding well.

No surprise either in seeing the three worst performers also as prime victims of the digital era. History books have to compete with the vast trove of material available on the web; that’s the Encyclopaedia Britannica syndrome, going out of print after 244 years of duty, demoted by the 11-year-old Wikipedia. Like it or not, most history books publishers will follow the same fate.

Similarly,Travel and Computer books are in direct competition with mostly free online services. Who will buy a “how-to” computer book today? There are plenty of video tutorials explaining how to replace a hard drive or how to struggle with Photoshop? And let’s not even mention the Travel segment with tons of guides, reviews, price comparators and transactions services. As for the language sections of the bookstore, again, a simple query in Google can help with spelling, translation and grammar… Even the precious Roget’s Thesaurus is online, and rather efficiently so. I’ll just venture French Canadians did Roget one better: A company called Druide publishes a suite of applications for PCs, tablets and smartphones called Antidote. It’s an unusually clever combination of dictionary, thesaurus, quotations, etymology and more. I wondered for a while about the name Antidote — until I realized Quebecois saw the product as an antidote to… English. An old struggle.

The main eBooks casualty is likely to be bookstores. In a city like New York, in the Fifties, about 330 bookstores were in business. Now they are down to 30 or even less, laments André Schiffrin, former head of Pantheon Books, in his recent book Words & Money. Countries like France or Germany have laws that protect independent bookstores: From Bordeaux to Berlin, citizens are thankful for finding warmer and more relevant recommendations than the algorithm-based suggestions provided by Amazon. But how long will it last?

frederic.filloux@mondaynote.com

Pro (Advertising) Choice

A couple of weeks ago, I came to a realization: I was becoming more and more reluctant to click on advertising banners because I feared I being digitally tailed for the next few months. When I mentioned this to friends, I noted that I was not alone. Everyone had their example of ads that, once clicked, become as sticky as the proverbial band aid. This could be the result of exploring a product (read my own experience testing an app), or occasional research on a subject… Your online behavior — queries you send, ads you click on — draws your marketing profile, enabling brands to deluge you with “targeted” ads. A shoe freak will be swamped by shoemakers ads, someone who intends to buy a car will be targeted by automakers and dealers. (I always wonder how the web page of someone afflicted with an embarrassing disease looked like…)

Once you’re caught in the behavioral targeting net, you’ll have a hard time cleaning up your surfing. I recently tested a utility for my computer — a poor quality product I quickly dumped — and ended up having to spend time removing the offending cookies with metaphorical tweezers. Now, I sacrifice a “polluted” browser (and a specific email account) which I use to click on ads, download products or marketing information, and do my best to keep my other browsers clean.

Why not flush the hundreds of cookies piled up inside my browsers, you might ask? Good question. In a file on my two computers, I keep almost 200 encrypted passwords, ranging from subscriptions to various publications, accounts to e-commerce sites or business online services. I don’t want to re-enter these codes each time I get rid of unwanted cookies. Hence the “dirty” browser.

The conclusion is obvious: behavioral advertising is backfiring. The more experienced users become, the more cautious they get in order to avoid aggressive tracking. For advertisers, this is the exact opposite of what they meant to achieve. And I take the trend will accelerate. Marketers have more sense of efficiency than of measure; they were quick to embrace these clever technologies without considering they might end up killing the golden goose. It is happening much earlier than anyone has anticipated.

The debate around the Do Not Track (DNT) system epitomizes this trend. The idea originated at the US Federal Trade Commission (FTC): it devised a piece of software embedded in a browser or an application, able to send a signal instructing a web site not to inject a tracking cookie in the user’s computer. After that, it is up to the website to comply or not. Mozilla quickly included the feature in its version 9.0 of Firefox, and Twitter followed.

Early june, Microsoft added fuel to the fire by announcing the DNT feature will be turned “on” as a default on its new Internet Explorer 10 browser set to work with Windows 8. This is by no means unimportant: the vast majority of users do not change default settings in their software. As a result, a sizable percentage of web surfers could end up automatically asking web sites to forgo any tracking. A potential catastrophe for the advertising industry: while most ads are purchases in bulk, at heavy discounts, the industry relies on behavioral targeting to increase the efficiency of ads — and of their resulting margins.

Intense lobbying on behalf the ad community ensued.

First, the definition issue, As viewed by the FTC :

An effective Do Not Track system should go beyond simply opting consumers out of receiving targeted advertisements; it should opt them out of collection of behavioral data for all purposes other than those that would be consistent with the context of the interaction.

Naturally, marketers are in favor of a much narrower definition, excluding the data collection process. In other words, OK for not targeting users, but their personal data must be ours.

In this story, Atlantic’s senior editor Alexis Madrigal makes the following point:

No one understands the industry’s definition because it deviates so far from the standard english definition of the word ‘track.’
Stanford’s Aleecia McDonald found that 61 percent of people expect that clicking a Do Not Track button should shut off *all* data collection. Only 7 percent of people expected that websites could collect the same data before and after clicking a ‘Do Not Track’ button. That is to say, 93 percent of people do not understand the industry’s definition of DNT.

Eventually, Microsoft had to backtrack under pressure from the Digital Advertising Alliance. The DAA is a one-year old body that defines itself as the “Self-regulatory program for online behavioral advertising”; it lines up all the major players in the business, including Google, Apple and Microsoft. The DAA fired a first shot by saying that the “on” default setting envisioned by Microsoft was going way beyond FTC’s definition as well as the W3C (World Wide Web Consortium)’s DNT recommendation. The DAA suggested DNT activation ought to be left to users — for instance, when they launch their browser for the first time. As a consequence, Microsoft’s IE10 featuring a DNT set to “on” as a ‘‘factory default’’ would be seen as “non-compliant” and the no-tracking signal sent to websites could be legally ignored.

The battle is just starting. It is unclear if Microsoft will fight the non-compliance issue and what kind of compromise will be reached. (The DAA’s final position will be disclosed in a few months.) In the meantime, digital kremlinologists will keep dissecting Microsoft true motives. After all, according to eMarketer, this year, in the US alone, the Redmond giant will make $700 million in advertising revenue:

This chart also clearly shows what’s at stake here. With DNT-as-a-default, Microsoft is obviously aiming Google and Facebook — and their higher advertising income. Both rely heavily on data-collection to serve relevant ads. It is even a crucial part of Facebook’s business model (see this previous Monday Note: Facebook’s Bet on Privacy) based on people giving up personal data in exchange for its service. A bet increasingly at risk.

frederic.filloux@mondaynote.com

Apps Features: Social vs. “Related”

Mobile application design is hard. For websites, we have well-established graphic rules. For PC screens, the tolerance for interface mishaps is fairly broad. Mobile apps are the  opposite: space is much scarcer, every pixel counts. Try shrinking a tablet app screen down an to a smartphone size: homothecy (linear reduction) rarely works. This is the reason why we often see fine iPad applications turn into flunked smartphone ones. It sometimes takes a while for a successful iPad app turn in to a well-adapted iPhone one: Flipboard, Zite and Bloomberg BusinessWeek were wise enough to take as much time as needed to roll-out great apps for the small screen.

When designers (and marketeers) perform user tests for a small screen app, they realize their design will have to adjust to many new circumstances and constraints. Reading time and general use conditions change substantially from a tablet to a smartphone: while the former is definitely a lean-back device, the latter will be used in many different ways, often including uncomfortable settings — I glance at my phone in a lobby, a waiting line, in the subway, etc. All this deserves thoughtful consideration when designing an application. The same applies to advertisers: they can’t expect to capture the same level of attention when moving from tablets to the smartphones.

With this in mind, I made a quick list of mandatory features for mobile applications.

Social vs. “Related”. Today’s hype leaves no other option but making an application as “social” as possible. This being the certitude du jour, allow me to think differently.

True, some apps are inherently social: when it comes to rating a product or a service, the “crowd factor” is critical. Beyond that, it should be a matter of personal choice — an antinomic notion to today’s the “Social” diktat. When you enroll into Spotify, the default setting is to share your musical taste with your Facebook friends and to suffer theirs in return. I personally can’t stand such obligation: I quickly dumped the application and cancelled my account.

The social idea’s biggest mistake is the belief in a universal and monolithic concept everyone is supposed to be willing to embrace with a similar degree of scope and enthusiasm. That’s a geeky, super-cartesian, Zuckerberg-esque view of society. Among my friends, some like opera (the singing, not the browser), others prefer heavy metal and I’m more into jazz tunes; some are tech-minded like me, others are more inclined towards literature. When it comes to sharing news, I tend to be naturally selective about the people I send a link to: I don’t want to swamp everyone with stuff they don’t care about. I might be wrong, but this is the way I see the social cyberspace: segmented and respectful of each other.

Where am I getting with this: When I read news online, I care about what is related (i.e. recommended by editors) as much as what is social (recommended by the crowd). Of course, Trending, is a good indicator of what’s hot. Here is a good example on TechCrunch iPad app, by any measure a thoughtfully designed one. Its Trending sidebar cleverly displays what’s hot and how it evolves:

Even better: when you dive into a story, the app will give you a better-focused “Trending” indicator on a particular company, in this example Buddy Media….

… will send you to the Crunch Base repository of people and companies:

TechCrunch’s social treatment is mostly Twitter-based. Subjects are connected to relevant tweets with the underlying story shown in a web view:

Related contents come in different flavors. Take the Bloomberg way shown in its remarkable BusinessWeek application. Companies mentioned in a stories can pop-up in a black sidebar drawn from the Bloomberg financial app.

Similarly, ProPublica’s application uses a lateral “drawer” to display related contents in an efficient way:

These features are by no means secondary. Providing related contents or a supplement of data, such as financial of biographical information, is the best proven way to retain users.

Finally, a word about graphics. Apple and the iPad have set the bar pretty high and very few apps takes fully advantage of their graphics power. One company rises way above the crowd: Roambi, in a class in itself when it comes to visualizing information. My take is, someday, most business sites will borrow from Roambi’s spectacular way of displaying graphics (part explanation of its design sophistication: the core of Roambi’s designers comes from the video game industry).

One last world about the ongoing debate between open web-apps and proprietary ones such as iOS or Android: The gap is narrowing. The FT.com, which pioneered the genre two years ago, made tremendous progress in its app. Periodically, a new release comes up with slight improvements in fluidity and ease of use. The iOS system and its software development kit remain a must for games and 3D intensive applications, but for news and data apps, HTML5 is getting closer.

One feature, though, is missing in most of these apps: the ability to use them offline. 3G coverage and cellular data transfers are more unstable than developers tend to believe; users should have more leeway in configuring their apps to download content in the background, ready for later offline use.

frederic.filloux@mondaynote.com

Facebook: The Collective Hallucination

Facebook’s bumpy IPO debut could signal the end of a collective hallucination. Most of it pertains to the company’s ability to deliver an effective advertising machine.

Pre-IPO numbers looked nice, especially when compared to Google at this critical stage of their respective business lives:

Based on such numbers, and on the prospect for a billion users by the end of 2012, everyone began to extrapolate and predict Facebook’s dominance of the global advertising market.

Until some cracks began to appear.

The first one was General Motors’ decision to pull its ads off Facebook. This was due to poor click-through performance compared to other ads vectors such as Google. No big deal in terms of revenue: according to Advertising Age, GM had spent a mere $10 million in FB ads and a total $30 million maintaining its presence on the social network. But Facebook watchers saw it a major red flag.

The next bad signal came during the roadshow, when Facebook issued a rather stern warning about its advertising performance among mobile users.

“We believe this increased usage of Facebook on mobile devices has contributed to the recent trend of our daily active users (DAUs) increasing more rapidly than the increase in the number of ads delivered.”

If Facebook can’t effectively monetize its mobile users, it is in serious trouble. Numbers compiled by ComScore are staggering: last March, the average American user spent 7hrs 21 minutes on mobile versions of Facebook (80% on applications, 20% on the mobile site). This represents a reach of more than 80% of mobile users and three times that of the next social media competitor (Twitter), see below:

(source : ComScore)

More broadly, Facebook experiences the unlimited supply of the internet in which users create inventory much faster than advertising can fill it. This trend is known to push ads prices further down as scarcity no longer contains them. The reason why the TV ad market is holding pretty well is its lasting ability to create a tension on prices thanks to the fixed numbers of ad slots available over a given period of time.

Unfortunately for its investors, in many ways, Facebook is not Google. First of all, it has no advertising “killer format ” comparable to Google’s AdWords. The search engine text ads check all the boxes that make a success: they are ultra-simple, efficient, supported by a scalable technology that makes them well-suited for the smallest advertisers as well as for the biggest ones; the system is almost friction-free thanks to an automated market place; and its efficiency doesn’t depend on the quality of creation (there is no room for that). One cent a time, Google churns its enormous revenue stream, without any competition in its field.

By contrast, Facebook’s ad system looks more traditional. For instance, it relies more on creativity than Google does. Although the term sounds a bit overstated considering the level of tactics Facebook uses to collect fans and raise “engagement” of any kind. For example, Tums, the anti-acid drug, developed a game encouraging users to throw virtual tomatoes at pictures of their friends. On a similar level of sophistication, while doing research for this column, I landed on the Facebook Studio Awards site showcasing the best ads and promotional campaigns. My vote goes to the French chicken producer Saint Sever, whose agency devised this elegantly uncomplicated concept: “1 ami = 1poulet” (one friend, one chicken):

If this is the kind of concept Facebook is proud to promote, it becomes a matter of concern for the company’s ARPU.

Speaking of Average Revenue Per User, last year, Facebook made $4.34 per user in overall advertising revenue. A closer look shows differences from one market to another: North America, the most valuable market, yielded $9.51 per user vs. $4.86 for the European market, $1.79 in Asia and only $1.42 for the rest of the world. Facebook’s problem lies exactly there: the most profitable markets are the most saturated ones while the potential for growth resides mostly in the low-yield tier. In the meantime, infrastructure costs are roughly identical: it costs the same to serve a page, or to synchronize a photo album located in Pennsylvania or in Kazakhstan (it could even cost more per user in remote countries, and some say that FB’s infrastructure running costs are likely to grow exponentially as more users generate more interactions between themselves).

Facebook might be tempted to mimic a rather questionable Google trait, that is “The Theory Of Everything”. Over the last years, we’ve seen Google jumping on almost everything (including Motorola’s mobile business), trying a large, confusing array of products and services in order to see what sticks on the wall. The end result is an impressive list of services that became very valuable to users (mail, maps, docs). But more than 90% of Google revenue still come from a single stream of business, search ads.

As for Facebook, we had a glimpse already with the Instagram acquisition (see a recent Monday Note), which looked more like a decision triggered by short-term agitation than by long-term strategic thought. We might see other moves like this as Mark Zuckerberg retains 57% of the voting shares and as the company sits on a big (more than $6 billion) pile of cash. Each month brings up a new business Facebook might be tempted to enter, from mobile phones, to search.

All ideas that fit Facebook’s vital need for growth.

frederic.filloux@mondaynote.com

Francois Hollande’s Start-down Nation

A forgettable election campaign just wrapped up: François Hollande is now the President of the French Republic. Time spent on foreign issues during last week’s one-on-one television debate mirrored the rest of the campaign: less than fifteen minutes in a 2hrs 50 minutes bout, one that left most viewers yawning. This campaign was petty, gallic-centered, oozing with demagoguery and completely devoid of great projects or ambitions for the country.

Mr. Hollande himself epitomizes this political flabbiness. He’s the default candidate. Last year, after Dominique Strauss-Kahn’s sexual implosion, Hollande kept running his tiny electoral diesel engine in low gear, bereft of grand ideas, unable to get into overdrive. He didn’t win because of his track record — he has no such thing. He was both a mediocre party leader and the weak manager of the poorest French department, which he left heavily indebted. Mr. Hollande didn’t win because he embodies any kind of grand aspirations either; other than getting into the Élysée palace, he has none. Instead, he portrays himself as a “normal guy” after — it’s also fair to mention — the hyper-kinetic, agitated, communication-obsessed, Nicolas Sarkozy.

Sure thing: with François Hollande, the country will rest; it will settle into gentle indolence as the rest of the world evolves and interacts. Most likely, the Euro zone crisis will heat up with a worsening situation in Spain where a quarter of the population — and half of the youngest citizens — are unemployed. Most likely also, ratings agencies will further downgrade French debt — since 1973, the country never had a balanced budget. Meanwhile, Mr. Hollande will fulfill his electoral promise of hiring 60,000 additional teachers; this while the country needs less teachers (there are less kids in front of them) but higher paid ones, along with higher respect and better working conditions. He’ll travel to Brussels and renegotiate the European Treaty, pitching the notion of growth (eureka!) against the “austerians“. On this, he might be right somehow (see last week’s Paul Krugman’s column in the New York Times titled Death of a Fairy Tale.)

All this doesn’t mean the Socialist presidency will be as dangerous as too many like to say. People making more than one million euros per year will indeed enter a 75% tax bracket: The new president claimed “[he] doesn’t like rich people” (a few years ago, he assigned a threshold of wealth to the equivalent of $60,000 a year). But vis-à-vis the much-bashed “Finance”, he’s likely to act as a pragmatist. A possible chief of staff for Hollande could  be Jean-Pierre Jouyet, currently chairman of the Financial Markets Authority and former minister of European affairs. There is no shortage of left-leaning talented people, and this middling leader is likely to surround himself wisely — on many counts, he can’t do worse than his predecessor.

France won’t fall from the cliff, nor will it shine brightly under the new regime.

And it won’t innovate either.

What made this campaign so depressing was both sides seemed to willfully ignore one of the most potent engines of the economy, that is innovation and a country’s ability to foster it. Both candidates seemed totally disconnected from critical challenges in which France is failing in every possible way.

Take higher education. The failure is unequivocal, regardless of political leanings. France might have about 80 universities, most of them second or third rate and producing mostly unemployable people. And if you dare a transatlantic comparison, you generate killer statistics. France’s budget for higher education and research is the equivalent of Harvard University’s endowment (€24 billion or $31 billion for French universities and public laboratories and $32 billion of cash reserves for Harvard). Overall, France’s spending per student is less than half of the US — and 15 times less if you compare to the Ivy League colleges. French faculty members, unions and politicians have made their best efforts to disconnect universities from the business world. They’ve been remarkably successful. As a result, Gallic colleges have become poorer, and largely unable to cope with the legions of students that land onto their benches, facing underpaid and unmotivated professors.

Of course, France has a different way to produce — and to reproduce — its elites. Two highways, actually: l’Ecole Nationale d’Administration (ENA) and l’Ecole Polytechnique. Mr. Hollande is an offspring of the first (so was his former partner, Ségolène Royal, the unlucky but picturesque 2007 presidential candidate.) As someone who grew inside this comfy seraglio and who traveled very little abroad, the new French president can’t envision an alternative to this trusted model for running the country. As for Polytechnique, it produces the top French engineers, a caste in itself, that has little to do with those graduating from top anglo-saxon colleges. The difference between a Polytechnique student and a Stanford one is the former will dream to manage, one day, a large industrial concern such as Thales (defense electronics) or the energy group Total, while the Stanford grad will want to see his/her name on a campus building — after a creating a successful business, needless to say. As the New York Times noted in a recent story about the return of class war,

Just under half of France’s 40 largest companies are run by graduates of just two schools: ENA,(..) and École Polytechnique (…). Together the schools produce only about 600 graduates a year. There are fewer than 6,000 ENA graduates alive today, compared with at least 160,000 Oxford alumni.

This doesn’t constitute the best soil for a start-up culture. And the venture capital activity is not likely to help either. In 2011, French VC funds invested €822 million in start-ups, a 21% drop vs. 2010. Even worse, 64% of these funds went to second or later rounds of financing, initial funding collected a mere 8% of the total. Not exactly a risk-prone attitude.

Again, international comparisons hurt. French VC invested last year about €13 or $16 per company and per inhabitant; that compares to $93 in the United Sates and more than $110 in Israel. Speaking of Israel, if we take into account the money flowing from abroad, the figures are even more staggering as VC funding per capita rose to $280 in 2011 according to IVC-KPMG data:

In 2011, 546 Israeli high-tech companies attracted $2.14billion from local and foreign venture investors, the highest amount in 11 years. This is almost 70 percent above the $1.26 billion raised by 391 companies in 2010 and 91 percent above the $1.12 billion raised in 2009.

In their excellent book Start-up Nation, the story of Israel economic Miracle, authors Dan Senor and Saul Singer also write:

Comparing absolute numbers, Israel — a country of just 7.1 million people –attracted close to $2 billion [in 2008] in venture capital, as much as flowed to the UK, Germany and France combined. (…) In addition to boasting the highest density of start-ups in te world (a total of 3,850, one for every 1,844 Israelis), more Israeli companies are listed on the Nasdaq [list here] than all companies from the entire European continent.

To complete this quote: about 250 companies originating from Israel had an IPO on the Nasdaq. Today 50 companies remain listed vs. 47 European companies including 3 French ones.

None of the above was mentioned, even remotely, during the French election campaign. Nicolas Sarkozy did very little about fostering innovation — he didn’t have a clue. As for François Hollande, the strongest part of its electorate (largely composed of teachers and other public servants) opposes any rapprochement between private sector and public higher education. And let’s not mention the underlying “ideology” of venture capital, carried interest, IPO’s, flexible employment rules, etc. Hollande’s supporters will also oppose any removal of cobwebs from the 102-year-old labor code that greatly complicates the management of companies employing 50 or more people. As a result, France has 2.4 times more companies with 49 employees than with 50, read this story in Bloomberg BusinessWeek.

This makes France a rather start-down nation. Nothing to celebrate.

frederic.filloux@mondaynote.com

Advertising: The trust Factor

The digital advertising equation is outlined in the Nielsen graph below. The Global Trust in Advertising survey released this month (summary on Nielsen site and PDF here) underlines one key finding: For the vast majority of digital users, trust lies first and foremost in recommendations and opinions from their peers. As for the bulk of formats found on web sites or on mobile (such as various flavors of display advertising), they fall to the bottom of the chart. Nielsen’s study, based on 26,000 respondents in 56 countries, was conducted in Q3 2011.

Here are the expanded results (click to enlarge):

By themselves, these figures provide the perfect explanation for the current state of the advertising industry and, more specifically, for the digital ads segment.

Then, superimposing the ad revenue structure of most news medias companies would show an alarmingly symmetry: these businesses derive most of their revenue, allocate most of their effort to the least trusted ad vectors: display banners of various forms (on web, mobile or social), online video ads, etc.

The survey also provides a grim view of what people trust: they put more of their faith in a branded website (58% positive), a brand sponsorship (47%) ad, or even a product placement in a TV series (40%) than in a display ad on a website or on mobile (33% each)!

Even worse is the general distrust of advertising: on this list of 19 ad vectors, only 5 are are trusted by 50% of the respondents.

Let’s focus on a few items:

Recommendation from people I know: Trusted: 92% Not Trusted: 8%
Consumer opinions posted online: Trusted: 70% Not Trusted: 30%
Problem is: traditional medias don’t own these two segments. Social networks and consumer websites do. It’s a key Facebook’s strength to have people engage in conversations around brands and products. (IMO: a pathetic waste of time). Interestingly enough, the social network environment doesn’t boost the despised banners that much: When served on a social network, banners gain a mere 3 percentage points (at 36%) against a plain website or a mobile context. This must be a matter of concern for Facebook’s revenue stream: its unparalleled ability to pinpoint a target doesn’t raise the level of trust.

Editorial content such as newspaper articles. Trusted: 58%, Not trusted: 42%
Not surprising, but worth a bit more thought. It pertains to the level of trust readers put in the medium of their choice — carbon or bits. As expected, a fair and balanced product review written by a non-corrupted journalist (every word in the sentence counts) will be trusted. That’s what I call the Consumer Reports syndrome. This organization deploys 100+ professionals testers — and no ads beyond the ones for its own paid-for services and extra publications. Among its enviable base of 7 million subscribers, half pay $6.95 a month (or, a much better deal, $30 dollars a year) to access ConsumerReports.org — this is good ARPU compared to other digital medias who only make a few bucks per year and per viewer in advertising revenue.

What does this mean for online outlets? They should consider beefing up the volume of product reviews, while preserving the reliability of their coverage. This also raises the question of the separation between journalism, advertorial and plain advertising. By no means should a publisher accept blurring the lines: beneficial on the short term but damaging on the long run. Having said this, when I see a growing number of anglo-saxons magazines making big money from high quality advertorials, I tend to believe online medias should consider sections of their websites or applications harboring such content. But two requirements need to be met: (again) no confusion whatsoever; and editorial standards for what will indeed carry commercial content, but in a well-designed, informative, visually attractive package. One important point to keep in mind: this type of service is typically out of reach for a Facebook, a Google or a Microsoft. But moving in such a direction requires unified thinking between publishers, the sales house (and the ad agencies they are dealing with) and the editorial team. A long way to go.

Ads served in search engine results:  Trusted: 40% Untrusted: 60%
Speaking of Google, here’s another interesting finding in the Nielsen survey: by and large, readers doesn’t trust search ads. To many viewers, text ads popping up on pages, on YouTube video or on emails, are seen as intrusive and irrelevant (to say the least: look at this hilarious site featuring inappropriate ad placements.) Still, search ads account for about 60% of online ad revenues. Why? Essentially because it provides a cheap, convenient, and totally disintermediated way of promoting a product. On this count, Google makes no mystery of its intention to vaporize the advertising middleman thanks to its superior technology.

The digital advertising party is just warming up. The business will continue its ongoing transformation. Currently, digital accounts for 16% of the global ad spending. It is likely to gain 10 more percentage points over the next five years. Not all markets nor products carry the same potential: According to the Financial Times, Unilever currently spends 35% of its US budget on digital, compared with 25% in Europe and only 4% in India. For news medias, the opportunity is that brands and agencies are still searching for the right formula. Brands face an incredibly complex challenge as they have to play with many dials at the same time: traditional ads, digital, web, mobile, apps, social, behavioral. And all are tightly intertwined, creating flurries of new metrics: ROI naturally, but also engagement, sentiment, feelings.

Like elsewhere in the digital world, the most successful players will be the genuine tinkerers. Software giant Adobe is said to spent 20% of its digital budget on experimental campaigns. They test, measure, adjust and iterate.

It is up to digital medias to go from passive to active in the quest for the right model. Their economics depend on it.

frederic.filloux@mondaynote.com

NYT Digital Lessons

The New York Times Company’s latest quarterly numbers contain a rich trove of data regarding the health of the digital news industry. Today, we’ll focus on the transition from traditional advertising to paywall strategies being implemented across the world. Paywall appear as a credible way to offset — alas too partially — the declining revenue from print operations.

First, the highlights.

(See NYTCO’s press release here and stock here. Unless otherwise stated, all figures are for Q1 2012 and comparisons are Q1 2012 vs. Q1 2011.)

  • Total Revenue is stable at $499.4 million.
  • Operating profit is down by 23% at $19.6 million. When excluding depreciation, amortization and (generous) severance packages, OP is up 9.4% at $57 million.
  • Print advertising for all properties and from all sources is down 8.1% at $238 million
  • Circulation revenue is up 9.7% at $227 million.
  • Digital subscriptions, launched just a year ago, reach 454,000. That’s a 16% growth vs. Q4 2011.
  • Digital advertising for the entire NYTCO (this includes NYTimes.com, BostonGlobe.com, Boston.com, About.com, etc) is down 10.3% to $71 million.
  • Such decrease is primarily due to About.com losing 24% of its ad revenue to $22.6 million, and 50% of its operating profit to $7 million. This online guide is entirely dependent on advertising.
  • But the real bad news is the decline in digital advertising for the NYT News Media Group  consisting mostly of the NYT and the Boston Globe. Revenue dropped by 2.3% to $48.5 million for the quarter.
  • Digital advertising accounts for 22.5% of the entire NYTCO ad revenue, and for 30% of the NYT News Media Group’s digital advertising revenue.

We can discern four trends:

#1:  Digital advertising is struggling, even for a major brand such as the New York Times.
Again the evolution :
FY 2010: +18%
FY 2011: +10%
Q1 2012 (Y/Y):  -2%

This confirms a much feared trend. By and large, in a news context, the performance of digital advertising is on the decline. All indicators are now flashing red: CPM (cost per thousand impressions), cost per click, volumes, yields, etc. The cause is well-known, and way more acute for digital than for print: ads and news contents do compete for the same eyeballs. The more attractive and eye-catching the content is, the lesser the ad yields. Behavioral advertising won’t change that much — at least for hard core, high value-added news environment.

This decline also announces a major shift in the way ads are sold. The advertising flow is likely to split: premium ads such as well-placed special packages will still be sold for high prices by in-house teams. But the bulk of the inventory will shift downward to bazaars in which gazillions of pageviews will be dumped into real-time exchanges supposed to optimize prices. The bad news: such schemes are likely to fuel deflationary trends for remnant (i.e. sub-premium) inventories. The good news: media organizations such as online news outlets or pure players are likely to join such marketplaces and perhaps gain an operating role of sorts — assuming they are smart enough to cooperate (I’ll address this in an upcoming column).

#2 Paywalls work. With roughly half a million paying subscribers, the NYTimes.com has captured the equivalent of 39% of its weekday print circulation of 1.3 million. In its financial statements, the Times doesn’t break down its revenue structure, but a significant part of the 13% increase in circulation revenue (print + digital) is attributable to digital subscriptions (the rest comes from the recent print price hike).
Estimates are difficult but here are some clues: on these 500,000 digital subs, it is estimated that 60% pay the basic $15/mo rate while 40% opt for the full $35 digital package. This would translate to digital subscribers contributing $34.5 million (18%) to the $190 million in NYT Media Group circulation revenue that appear in its quarterly statement. 18% is not that bad for a paywall that is barely one year old (even though this estimated revenue doesn’t reflect the cost of the NYTimes’ massive promotions for its paywall program). But again, compared to the $48 million of digital advertising, it is significant.

#3 A warning to paywall dreamers: some restrictions apply. In order to be successful, a digital subscription must check the following boxes:
Own a sizable share of a given (and preferably solvent) segment of the population. In other words: start from a large built-in audience. Globally, the New York Times has about 34 million unique visitors per month – a large pool for conversions to the paywall.
Don’t expect a paywall to work for a small site or a niche product — unless it is a reference for its community. Even then, in spite of its reference status in New England, the Boston Globe shows a mere 18,000 paid-for digital subscribers.
– Allow time to grow the subscriber base. A paywall strategy must spread over several years. The free audience first has to be converted into registered users able to be thoroughly data-mined; then the paywall will be tightened with less and less articles available for free (the NYT recently lowered its threshold from 20 to 10 free articles); the entire process will take at least two to four years, depending on where you start from.
– Carefully manage porosity. That’s why some people refer to a “semi-permeable membrane” (see the interesting conversation between Clay Shirky and NYT’s Digital manager Denise Warren on NPR last January). While it is tightening its paywall, the NYT leaves willingly plenty of free access to its content: if you land its site from a search engine, from Facebook, Twitter, or from a blog, no limit applies (same for the FT.com, actually). Such tactic has two virtues: it doesn’t affect natural referencing and incoming traffic from search engines (which could weigh as much as 30-40% of the audience), and the brand remains exposed to many — such as social networks users.
– Quality is non-negotiable. A successful paywall requires exclusive, unique, authoritative, high-quality content. A paywall isn’t the right solution for streams of “commodity news” or user-generated contents. It won’t work for the Huffington Post. Despite its enormous audience, the HuffPo’s embryonic original content won’t do much to alter its “Left wing Fox News” positioning (Even though the HuffPo managed to score a Pulitzer Prize for National reporting for its remarkable Beyond The Battlefield series.)

#4 Print is still alive. While print advertising is drying up, the share of circulation revenue keeps rising (in relative terms.) The good news: price hikes don’t seem to matter: the recent increase to $2.50 had no effect on sales. Actually, the Times uses its weekend edition (priced at $5.00) to channel digital subscriptions by providing the best deal of its complex rate card. Which leads to two conclusions: a sizable reservoir of readers is ready to pay for quality-on-paper at almost any price (see a previous Monday Note Cracking the Paywall); and commercially strong weekend editions can be a potent vector for digital subscriptions.

Print and digital strategies are more intertwined than ever.

frederic.filloux@mondaynote.com

Facebook in Frantic Mode

Facebook’s acquisition of Instagram — for one billion dollars — tells a lot about Mark Zuckerberg’s state of mind. Which is at least as interesting as other business considerations and was best captured by cartoonist Ingram Pinn in last week’s Financial Times comic. To illustrate John Gapper’s excellent Facebook is scared of the Internet column, Ingram Pinn draws an agitated Mark Zuckerberg frantically walking through a hatchery, collecting just hatched startup-chicks as fast as he can, while, in the background, AOL and Yahoo collect older chickens in larger carts.

In last week’s Monday Note, I hinted that I’d never put my savings in Facebook’s stock. (For that matter, I see writing on business and owning stocks as incompatible). When I read the news of the Instagram acquisition, I wondered: Imagine Facebook already trading on the Nasdaq; how would the market react? Would analysts and pundits send the stock upward, praising Zuckerberg’s swiftness at securing FB’s position? Or, to the contrary, would someone loudly complain: What? Did Facebook just burn the entire 2011 free cash-flow to buy an app with no revenue in sight, and manned by a dozen of geeks? Is this a red-flag symptom of Zuckerberg’s mental state?

Four things come to mind.

1/ Because he retains 57% of Facebook voting rights, Zuckerberg rules its board and can make any decision in a blink of an eye, no debate allowed. This can be a great asset in Silicon Valley’s high speed tempo, or it can stir up shoot-from-the-hip impulsiveness.

2/ Facebook’s founder attitude reminds one of Bill Gates during Microsoft’s heydays: no crack allowed in the wall of its dominance. The smallest threat must be eliminated at any cost. Where Microsoft used legally dubious tactics, Facebook unsheathes its wallet and fires a billion dollars round. In the startup world, this will have two side effects: For one, Facebook is likely to become the exit of choice Google once was. Two, the size of the Instagram transaction (some of it in stock) is likely to act as a beacon for any startup harvesting users by the millions. It sets an inflationary precedent.

3/ By opting for such a deal, Facebook’s management reveals its own feelings of insecurity. It might sounds crazy for a company approaching the billion users mark and providing an array of services that became a substitute to the internet’s basic functions. But, with this transaction, the ultra-dominant social network acted like an elephant scared of a mice. Instagram has 35 million users? Fine. But how many are using the service more than occasionally? Half of it? How many are likely to switch overnight to a better app? Most likely many will. Especially since Instagram is not a community per se, but a gateway to larger ones such as Twitter and Facebook.

4/ From a feature-set perspective, Facebook might find itself in a quandary. Kevin Systrom and Mike Krieger designed the ultimate stripped-down application: a bunch of filters and a few basic sharing features. That’s it. It is both Instagram strength and main weakness. Such simplicity is easy to replicate. At the same time, if Facebook-Instagram wants to raise the feature-set bar, it might lose some of its users base and find itself competing with much better photo-sharing applications already populating Apple or Android app stores.
To put it differently, Facebook photo-sharing model had been leaking for a while. Zuckerberg just put  a serious plug on it, but other holes will appear. A couple of questions in passing: Will Facebook continue to accept and encourage loads of third parties photo-sharing apps that connect to its network? Some are excellent — starting with Apple’s iPhoto, especially the iDevice version that will always benefit of an optimized hardware/software integration. How does Facebook plan to deal with that? And if it chooses to grant some level of exclusivity to the Instagram app, how will the audience react (especially when you read comments saying “We liked IG because it wasn’t FB”)?

Lastly, the bubble question. Again, three things.

First, let’s be fair. If indeed there is a new internet bubble, Facebook isn’t the only player to fuel it; investors who lined up at Instagram’s doorstep did it too. A few days before the deal, IG raised $50 million at a half billion valuation; Zuckerberg snatched the company by simply doubling the bet.

Two, comparing the FB/IG deal to Google’s in 2006 acquisition of YouTube for $1.65 billion doesn’t fly either. From the outset, everyone knew internet video was destined to be huge; it was a medium of choice to carry advertising. Therefore, the takeover by Google’s fantastic ad-machine was likely to yield great results. YouTube became a natural extension of Google services — just look at how competing services such as DailyMotion in France, or Vimeo are doing without the ad rocket-engine.

Three, the metrics used in an attempt to relativize the deal are dubious at best. Instagram had no monetization strategy–other that a lottery-like exit. This says applying any kind of cost per user ($33 for the theory in vogue) is bogus. Being unable to project any sustainable revenue mechanism makes such a valuation process completely pointless. In Instagram’s case, the only way to come up with a price tag was guessing the amount of money a small group of suitors–Facebook, Google and Twitter–might be willing to cough up for Instagram’s eyeballs.

If this deal shows one thing, it is the frenzied, cutthroat competition these three players are now locked in. Mark Zuckerberg is not through with collecting hatching eggs. He won’t be alone either.

frederic.filloux@mondaynote.com

Facebook’s Bet on Privacy

Would you buy Facebook shares? A few weeks ahead of its mammoth IPO, millions of people probably dream of getting a slice of it. Spreadsheet jockeys have done their job and demonstrated with unanimous conclusiveness that, indeed, Facebook deserves its expected $100 billion valuation — or that gravity’s law will inevitably apply.

Facebook numbers are both fascinating and frightening. The social network will pass the 1 billion members mark this year and the capillarity of its services is creating an alternate internet before our very eyes. It has already become a credible substitute for email; it soon will be the dominant news channel for millions. At the same time, it is hugely profitable: Facebook’s margin reaches 62% and its $3.5 billion cash pile will allow occasional mistakes or, if you prefer, bold experiments.

Then, what could go wrong for the ultra-dominant digital rhizome? Two things: its contempt of privacy and Wall Street frothy expectations.

Two years ago, I interviewed David Kirkpatrick for Le Monde Magazine. He’s the author of the Facebook Effect, a book that remains a must-read if you want to understand the company and its founder. In our conversation, he described a Mark Zuckerberg perfectly aware of the sensitivity of privacy issues, but at the same time deeply convinced social norms would evolve towards nearly-total transparency. According to Kirkpatrick, Zuckerberg felt that, as long as users where given the proper tools to control it, privacy should not be an issue for his empire’s future. Put another way, Zuckerberg deeply believes in Facebook’s Grand Bargain: its core followers will accept full openness as a default setting and trade personal data in exchange of its features-rich service. He actually lived by this belief. And monetized it brilliantly.

Facebook’s most valuable currency is not the “Credits” used in its games, but its huge trove of consumer data.

The efficiency of this system comes form the “platform effect”, from Facebook’s federation of millions of sites that embed the “Like” button or allow their own users to register with their Facebook ID. Trying to protect one’s privacy while using Facebook is a hard Protean task as the company constantly changes its rules. Applications hosted by Facebook only make things worse as user personal data are allowed to leak to third party developers in a sneaky and overly abundant ways.

Last week, as a part of its remarkable What They Know series, the Wall Street Journal published a compelling story titled Selling You on Facebook:

A Wall Street Journal examination of 100 of the most popular Facebook apps found that some seek the email addresses, current location and sexual preference, among other details, not only of app users but also of their Facebook friends. One Yahoo service powered by Facebook requests access to a person’s religious and political leanings as a condition for using it. The popular Skype service for making online phone calls seeks the Facebook photos and birthdays of its users and their friends.

You might ask: What’s the connection between these privacy concerns and the upcoming IPO? Well, Facebook derives most of its revenue from advertising. And said advertising revenue stems from its ability to profile its users like no one else in the business. Still, in spite of its ability to serve an ad targeted to a South Texas single mother who likes Bob Dylan and Taco Bell, Facebook yields little revenue per capita. Where Yahoo makes $7 per user and per year and Google $30, Facebook’s ARPU actually amounts to a mere $4.39.

A further problem: Facebook does so after saturating its most solvent markets.

Now, let’s turn to Wall Street expectations. A $100 billion valuation would mean Facebook being traded at 27 times its 2011 revenue. For comparison, Google, Apple and Microsoft, all highly profitable, are valued between 4 or 5 times their respective revenue for their last fiscal year.

Hence the math: In a recent story published in Fast Company, Farhad Manjoo quotes a Dartmouth finance professor who said “[to justify Wall Street expectations] Facebook will need to see $70 billion in annual revenue by 2021, up from just $3.7 billion in 2011″, which translates into a 25% to 30% growth over the next decade… in the context of an advertising market growing at 4% per year, as Manjoo points out.

To meet these goals–even by going way beyond the one billion members mark–Facebook will have to extract more bucks from each one of its users. This means making an even better use of the data users traded for services.

This brings us to the biggest risk for the Facebook economics. If Facebook doesn’t play the privacy game well, two things are likely to happen. One, members will pressure the social network to limit its use (meaning the sale) of user data. Two, legislators will enter the fray. We already see early signs of political challenges with movements such as the Do Not Track initiative, one that is laying the ground work for legislations all over the world. This concern was reflected in the Risk Factor section of Facebook’s S-1 filing :

Our business is subject to complex and evolving U.S. and foreign laws and regulations regarding privacy, data protection, and other matters. Many of these laws and regulations are subject to change and uncertain interpretation, and could result in claims, changes to our business practices, increased cost of operations, or declines in user growth or engagement, or otherwise harm our business.

Facebook’s future relies in great part on its ability to wisely adjust the privacy dials. Even at the expense of its shareholders’ dreams.

frederic.filloux@mondaynote.com

Culture Shift: User To Client

Fifteen years ago, Louis Gallois, the SNCF (French Railways) chairman decided to change the company’s lexicon: passengers were to be referred to as “customers” instead of the old bureaucratese “users” (in French: “clients” vs. “usagers”). The intent was to convey notions of choice and consideration for the rider. This being France, the edict led to convoluted debates. The upper management old guard held the company was on its way to betraying its traditional mission of service public. Unions—notoriously opposed to any forms of competition threatening their fiefdoms—saw the new word as a portent of evil mercantile designs. In Louis Gallois’s mind, a clientele should not be seen as captive herd but ought be shown respect and empathy. It took more than a decade to see the French railway system become more customer-oriented. The French Postal service underwent a similar transformation—largely under pressure from internet-based services. Today, when compared to most other countries, these companies have become good performers.

Back to my media beat, you see where I’m going: The transforming media industry is still stuck into a user’s culture. Media companies still believe this: One way or another, they own their readers (or viewers and listeners). Of course, this belief is not evenly shared among different corporate layers. In the C-suite, the comfy old view is long gone as numbers confirm, quarter after quarter, the industry’s slump. Most executives share a sense of vital urgency. But the deeper you dive into those companies, the more you see complacency still lurking.

As long as the old media culture still dominates and resists change, better business models won’t be able to gain traction.

It all boils down to a simple market place evolution.

In the pre-internet era, the media sector lived by its own rules: a captive audience left with no other choice but a bunch of well-entrenched media outlets. At the time, these companies didn’t feel the need to probe their audiences, let alone to market to them. People were listening, viewing, and reading, roughly at the same rate, year in and year out. Editors and publishers felt immune to any form of challenge. Newsrooms were a great place to be, filled with witty, smart people, most of them notoriously unproductive, but great to hang out with, caring very little care for the user’s state of mind.

Then, the digital wave unfurled. With it came a new business culture, completely antinomic to the legacy media’s thinking. At first, the tech/startup way of doing things was dismissed as a freakishly geeky and completely inapplicable to media organizations.

Then the two spheres—the new entrepreneurial culture and the old one— got closer and closer and began to intersect. The overlapping zone was, precisely, digital information. It began in chaotic but participatory (massively) and profuse ways. This led to the rise of “commodity news”—whose value evaporated in the process—at the expense of the original (and traditional) news sources that were slow to understand the scope of the upheaval. This put a brutal end to the widespread old complacency.

As the user morphs into a customer, s/he becomes more demanding of its media provider. There is a reason for that shift: a magazine subscriber is also an Amazon patron and s/he now expects the same level of service. Instead, for most magazines, it still takes 3-6 weeks for a monthly print subscription to start.

Today, the media industry must change its reference system. Every single day, traditional-media-in-transformation collides with companies (pure players, aggregators, portals, search engines, mobile applications, retailers, distributors) built on very different, opposed sometimes, values and principles. As a result, the competition on products (and audiences) leads to a competition on the processes of building and marketing these products.

This can be summed up to three notions.

The Customer (again).  He (she) is no longer a well-defined monolithic individual. Consider the structure of a digital audience: news consumption is scattered all over the day with different size and shapes. This should impact the way news is packaged. Most newsrooms are currently unable to adapt to the time-sensitive diversity that has become expected. Too many newsrooms don’t understand their output should be reformatted, re-edited, for different uses, at different times of the day, on different devices.

Competition / Speed => Leading the pack. The media business is now intensively competitive. Newsrooms should be obsessed with beating the competition in every possible way, exactly in the same fashion a tech company is constantly rolling-out new features for an application or a service. Unfortunately for the slowest and the weakest, the media industry is migrating to a “winner-takes-all” system, with very little oxygen left to the lower tier.

Responsibility / Empowerment / Focus => Better Execution. This implies two moves: First, a complete overhaul of the HR culture. The old media culture is plagued by poor accountability and dilution of responsibility. It’s time to shift to one project (or one segment of the business) = one Direct Responsible Individual, meaning true delegation, a clear mission, and the sanction (positive/negative) that goes with it. Two, it involves a change in the compensation structure, until then dominated by guild-management negotiated agreements that abhor genuine meritocracy. Again, the technosphere teaches us the benefits of the opposite: a human management system able to attract, retain and promote talented people. The combination of responsibility and reward (not only financial) is a non-negotiable requirement for better execution.

Before going back to spreadsheets and corporate dashboards, all the boxes above must be checked. Vaste programme.

frederic.filloux@mondaynote.com