Do Tumblr’s Audience Analytics Support its $1.1B Valuation?

Why is Yahoo paying $1.1 billion for a blogging platform most “grownups” have never heard of? The media has been busy postulating:

  • The New York Times suggests Yahoo “is seeking to make up for years of missing out on the growing use of social networks.”
  • “If Yahoo buys Tumblr, it’ll be a means of absorbing a throbbing, vibrant mass of teens and cool youths” writes VallyWag.
  • “117 million unique users world-wide in March…up from around 58 million a year ago. The site is among a number of fast-growing startups,”* says the WSJ, citing statistics from comScore, a measurement firm.
  • Trade publication AdAge adds that “the deal checks many boxes: social, mobile, content and buzzed-about native ads”

Is Tumblr really the “throbbing,” “fast-growing” and “social” solution to Yahoo’s languid business performance? Let’s take a look at some publicly available data to determine whether it supports these claims. This is not an in-depth financial analysis, but rather a broad look at some of the performance metrics related to audience and traffic, which are key drivers of revenue for web companies.

How cool is Tumblr’s audience?

If cool is defined by young users, Tumblr is at least halfway there: According to digital measurement firm Quantcast, 50% of Tumblr users are 24 and under. That’s a 20% point leg up on Yahoo, whose under-25 audience comprises less than 30% of total visitors. But while young, Tumblr may not be the “vibrant mass of teens” that VallyWag suggests: Quantcast says less than 20% of Tumblr users are under 18, a lower share than Facebook. Tumblr indexes highest among college-aged up to 35-year-old users. Think young adults who might be on MySpace were it not so passé.

Diversity might be considered another cool factor: Tumblr is more ethnically diverse than either Facebook or Yahoo, indexing especially strongly among Asian and Latino users. Just over 60% of its users are Caucasian — a reflection of where American will be rather than where we are now.  And importantly, these users are more likely to be college educated than either Facebook or Yahoo users. But, likely a function of life stage, Tumblr users are low income: Over 60% of Tumblr households earn less than $50k/year, significantly lower than Facebook and Yahoo users. If there is  value for advertisers in this audience, it may not be in the current buying power but instead future potential to shape society’s tastes.

Comparing Demographics: Tumblr, Yahoo and Facebook 

tumblr demographics vs. facebook and yahoo

 (Quantcast, April 2013)

How does Tumblr perform?

Coolness itself is not enough to sustain a $1.1 billion valuation.  What counts as cool is often ephemeral, especially among younger demographic cohorts who are changing their beliefs, attitudes, friends and brand affinities month-to-month.  The platform has to perform for Yahoo. Let’s examine some metrics that may indicate how sustainable it will be.

Tumblr is growing, largely due to increased adoption outside the USA. International visitors increased 35% year over year, while US-based visitors declined 2% for a net gain of 19%, according to Quantcast, which measures Tumblr directly with java script tags. This is a far cry from the 100% growth implied the WSJ above. The reason: WSJ used comScore numbers, which are typically respected as authoritative. But due to a technicality* they are not in this case.

Despite the audience growth, Quantcast data suggests Tumblr’s page views are down over last year. So while more users are visiting the site, they appear to be less engaged, viewing fewer pages every time they come. It’s possible this decline is due to design changes which make navigation simpler, requiring fewer clicks. There is not enough data available to know the true driver of the page view deterioration.

Still other data points suggest engagement is strong. Visitor loyalty, as measured by average visits per person every month, has increased modestly YOY within the USA. This indicates there is a cohort of increasingly engaged enthusiasts, even while slightly fewer people are visiting.  Although at 4.5 visits per month, Tumblr significantly lags behind Facebook’s repeat rate of 30+ visits per person. This may be an argument that Tumblr is a fundamentally different kind of social network, one that has a naturally lower level of user engagement.

Tumblr’s Key Analytics Vital Signs

tumblr traffic trending

 (Quantcast, April 2013)

While traffic metrics give mixed signals, the financials are less open to interpretation: In 2012, Tumblr made just $13M. That translates to revenues of just $0.11 per monthly unique visitor. The top line is expected to be significantly better this year, about ten times higher than 2012, thanks to the introduction of new ad products. That puts early revenue per monthly visitor at $0.72 – but it’s still a far cry from Facebook’s estimated $6 annual revenue per monthly visitor. It does in theory, however, justify the valuation of $1 billion. Companies typically sell for about 10x revenue.

Estimating Tumblr’s Valuation 

tumblr valuation estimates vs. facebook

But a 10x valuation assumes sustainability for the long term. While Tumblr undeniably injects aging Yahoo with a “pulsing” mass of expressive young things,it remains to be seen how this translates into tangible business results that last.

*comScore started directly tagging in April 2013. had previously only been tracked by comScore’s panel data. Direct tagging generally inflates visitor counts, rendering year over year comparisons meaningless.


The Economics Behind Does a Contributor Model Work?



The online publishing business is running on an accelerating treadmill: more and more output is required to stay in place financially as competition grows and ad rates remain low. The business mantra at most web media companies is get bigger, faster, and without paying for it. There are three popular grow-without-incurring-new-cost approaches:

  • Aggregation: Summarize other sources and link to them (think Huffington Post)
  • SEO (search engine optimization): Create content that lives on in perpetuity in search engines (think
  • Free/cheap outside contributors: Pay outside contributors little to nothing for privilege to write original content for your brand (

At a time when SEO and aggregation are all the rage among publishers, is an interesting example of a company that has fully invested in the outside contributor business. In 2010, Forbes media bought True/Slant, a platform that makes it efficient to manage lots of freelance writers – in effect pinning their entire online strategy on original content. “Ours is not a cynical effort to ride the Google wave [nor is it] about bending or breaking the rules of fair use to build community around content created by others,” wrote product director Lewis DVorkin on his blog.

As of this writing, claims to have nearly 850 contributors on its newly designed website that integrates True/Slant’s technology. This number includes 440 outside contributors and a full time staff of roughly 400. So how is this model working for them? In a recent blog post, Forbes’ product director offered some interesting insight into the performance of these 850 contributors in June:

Performance of Writers, June 2011

readers per writer on

The chart shows that a small minority of 25 writers—just under 3% of the total—attracted more than 100,000 readers each. The vast majority of contributors attracted less than 20,000 readers a piece in June. These numbers would certainly vary month-to-month, but 850 is a large enough sample size to assume a certain consistency over time.

To understand how this translates into business results, we need to make a few educated guesses beyond what Forbes publically exposed above. These include: 1) estimate the total unique visitors per segment, 2) the ad impressions generated 3) and the average advertising rate per 1000 impressions.

First, I’ve estimated the total number of visitors for each audience level using assumed averages per writer in June (for example, I assume the 100K+ group averaged 200,000 readers each):

unique visitors estimates at

The total audience of 13 million corresponds roughly to 3rd party analytics data for, so we can assume these numbers are in the right ballpark.  Next let’s estimate the monetary value of each segment.

Based on public analytics data from publishers similar to Forbes, I’ve assumed each visitor averages 3.5 pages per month, and each page on serves two ad impressions. So by multiplying the total visitors above by 3.5 page views/visitor and 2 ad impressions/page, we get total ad impressions. At an estimated rate of $10 per 1000 ad impressions (eCPM), the following revenue figures emerge:

estimated online revenue, june 2011

You can see that the top 25 writers averaged $14,000 in June, but the vast majority—86% of the contributors—only generated $350 each.  Factor in overhead costs, and the earnings per writer might be less than half of these numbers. This means the top writers at could make $7,000 a month, or $84,000/year. But this assumes that they consistently perform in the top 3% of writers all year, an unlikely feat.

To make this model work, Forbes would only be able to pay around 100 of these 850 writers full time salaries. This would mean gutting their existing newsroom of around 400 while expecting the same quality of content from unpaid/very cheap contributors. Such actions could in turn damage their brand and reduce advertiser interest. This is why publishers like Forbes, even with aggressive online strategies, still cling tightly to their aging print cash cows.

Predicting the NYT’s Paywall Success with Web Analytics


The internet has exploded with commentary on the New York Times paywall. Over 600 articles penned in just the past week, according to Google News.  While there’s no dearth of insightful opinion, it’s often missing an important component: supporting analytics data. In this post, I use hypothetical web analytics numbers to explain some of the fundamental paywall questions, including:

·How will the NYT’s overall web traffic be affected

·To what extent will paywall the cannibalize existing advertising revenue

·The net effect on the NYT’s digital revenue

Because the NYT makes limited web data public—we know they have roughly 30 million monthly US visitors but little more— it’s very difficult to answer these questions with precision. But with a little detective work, we can deduce some probable data. The key question we must answer is simple: How loyal are readers? This is critical because the metered paywall hinges on usage frequency.

Thankfully some publishers, such as The Atlantic and The Daily Beast, share their web data publicly. From these sites and others, it’s possible to piece together an educated guess about the NYT’s audience analytics. Based on my research, I concluded that roughly 25 to 35% of top tier news website visitors are loyalists, visiting either daily or weekly. These  loyal users as highly engaged, making up about 55-65% of the visits and a vast majority of page views. From a traffic perspective, this is this is an important distinction.

The NYT’s editor Bill Keller states:

“Those who mainly come to the website via search engines or links from blogs, and those who only come sporadically — in short, the bulk of our traffic — may never be asked to pay at all.”

Notice Keller does not define “traffic.” While it is true a majority of users may never see the wall,  the majority of page views are generated by a smaller, loyal subset of the unique users. So if traffic is defined as pages views, then the bulk of it could be impacted by the wall.

Based on my research of properties competitive with the NYT, I’ve estimated the following web analytics metrics for reader segments (US audience only):


Visits per Month

Monthly visits

Pages per Visit

Monthly pages viewed

Daily users






Weekly users






Monthly users






Passers by*










*Only 1/3 of these users visit in a month

The estimated data shows frequent users are critical to the NYT’s web activity: less than 10% of users are daily readers, but they generate 72% of the monthly page views under this model. Because the paywall affects these daily readers most, it seriously jeopardizes page view traffic – and by extension, revenue.

Estimating ad revenue

Most online adverting is sold as impressions, meaning revenue is directly tied to page views (the metric we calculated above). Rates vary considerably for news sites, from around $2 per thousands ad impressions up to $30 if it’s a valuable audience. For a general interest, quality publisher like the NYT, we can assume the rate is $8 for every 1000 ad impressions. Each page view usually contains two sold impressions. With this, we can calculate how much ad revenue each audience segment generates:

Estimated Monthly Ad Revenue – No Paywall

Monthly pages viewed

Impressions (2 per page)


Monthly Ad Revenue

Daily users





Weekly users





Monthly users





Passers by









(Note: the NYT does not break out revenue for It reports a total figure that includes and other properties it owns, inflating the numbers you may see reported online. That’s why these numbers are estimated.)

Loyalists generate the vast majority of the NYT’s revenue in this model. Out of $8.3 million, the most loyal 8% of users generate $6 million. Looking at it another way, passers by are worth about a penny while loyal visitors generate about $2.40 per month. Because the paywall impacts loyal users only, it will disproportionately impact ad revenue.

Putting up the wall

After you visit 20 articles in a month, the NYT is going to ask you to pay up. There are a lot of ways around this: social links, certain top news stories, advertiser supported subscription plans, and if you’re print subscriber, it is free. We know the vast majority of sign ups will come from a subset of the daily user segment because they’re most likely to hit the wall.

After looking at subscriber sign up projections from a variety of online sources—some higher, some lower— I’ve estimated subscription rates in context of the analytics data projections in the below chart. My revenue estimate assumes an average plan value of $20/month (prices range from $15 to $35 every four weeks).

Estimated Paywall Subscription Revenue


Sign up rate

Paying users

Monthly Revenue ($20/subscriber)

Daily users





Weekly users





Monthly users





Passers by









Although just 313,000 users pay in this model, they generate as much ad revenue as over 20 million visitors. These are clearly very different business models, an ad model focused on volume and a subscription model on a small group of loyalists. The question is: how well will they work side by side?

Because of their usage volume, all 2,500,000 daily users will hit wall. Using the subscription rates above, this leaves us with a loss of roughly 1,800,000 daily users out of an original 2,500,000. We will charitably assume that all these unwilling-to-pay daily users become weekly users. The weekly and monthly users who subscribe may become more frequent users, but this model assumes the amount to be immaterial.With this in mind, here’s a rough estimate of ad revenue impact:

Estimated Ad Revenue Under Paywall


Monthly pages

Ad Revenue Under Paywall

% Change

Daily users





Weekly users





Monthly users





Passers by










After the paywall goes up, stands to lose 47% of its online advertising dollars according to this model. The majority of the attrition would come from loyal daily users, while less frequent visitor segments could see increases as the daily users visit less.

And finally, the most important question: what is the net impact on total online revenue?

Total Digital Revenue Under Paywall

Ad Revenue under paywall

Subscription revenue under paywall


Change vs. ad-only model

Daily users





Weekly users





Monthly users





Passers by










Summing up the ad revenue under the paywall with the subscription revenue, we get the total in the below chart, $10.6 million. This hybrid ad/paywall approach generates 28% more revenue than the ad-only model. While they lost nearly half of the online ad revenue, it’s more than made up for my digital subscriptions. Interestingly, the largest percentage revenue gains come from less loyal users. This is due to the severe decline in ad revenue among the daily visitors, and the resulting growth in weekly and monthly user ad dollars and subscription revenue.

The purpose of this post is not to show that the paywall is worth it, but rather to simply highlight the importance of web analytics as a decision support tool.  Judging by the failures of several newspaper paywalls, many publishers do not pay enough attention to their analytics data. For the sake of quality journalism, I hope the NYT does.

Of course analytics alone will not predict success. There are many other factors that could push these numbers into the red or higher into the black, including:

  • Advertisers may pay more to reach the subscribers, but  pay less to reach the non-subscribers
  • If the ad inventory is only 50% sold out on average, then a loss of page views from the wall will not have nearly the impact on ad revenue as laid out in my model
  • The NYT may develop new advertising products that are less dependent on page views
  • Other major news sources could adopt this wall model, which might encourage more people to pay but also possibly result in price wars
  • Long term impact from losing the majority of daily readers on the NYT brand as “the world’s newspaper”

Button line: It’s a big risk for the NYT, but if they are focused on their web analytics data, I think they should be able to adjust where is needed to make it work. The key is to pay attention to loyal users, who are worth many times more than casual browsers.

Agree or disagree with my model and analysis? Let me know what you think.

Why walled gardens will wilt


Pretend you publish a magazine. A recent circulation audit says 4 out of 5 of your readers are grazers – they flip through a few pages in dentist waiting rooms, airport gift shops, and hotel lobbies. Your advertisers, hungry for loyal, engaged readers, start to flee. To preserve the business, how would you respond? If you can’t afford to invest in improving the product, you might make a last ditch effort and whip up some full page ads to convince those infrequent browsers to subscribe.

Sadly, this is how many publishers are reacting to the internet. And given the numbers, it makes sense. Let’s examine Time magazine: According to analytics firm Quantcast, 80% of’s visitors last month were passers by – entering by happenstance, from links on Google, social media or other sites. Because advertisers don’t value these kinds of fly-by readers, much of’s ad inventory is used to sell print subscriptions.

time magazine house ad

Time is hardly alone. An AdAge article from 2010, “Web’s Big Boost to Magazines? Selling Print Subscriptions,” aptly sums up the strategic thinking of many publishers. In an interview last May, The New Yorker’s web editor, Blake Eskin, put it bluntly: “The most immediate business goal of all Condé Nast websites is to generate print subscriptions.” While we might add tablet apps to this now, it’s the same logic: the open web is a marketing vehicle for walled gardens, where ostensibly, loyal advertiser-friendly readers abound.

But there’s a problem with this approach: in 2011, it’s not clear walled gardens are more effective at building the high quality, loyal readers advertisers covet than the open web. According to data obtained by Women’s Wear Daily, iPad magazine purchases are dropping: Vanity Fair iPad app sales fell from 10,500 in August to 8,700 digital editions in November. It remains to be seen whether Apples new subscription plan will help boost these numbers appreciably.

Using the web as a tool to convert users into print or app subscribers is smart in the short term. But as competition from the open web devises new ways to siphon off your readers, you’ll be fighting an uphill battle. To make matters worse, advertising messages should become ever more effective on the open properties that have larger, loyal audiences. Let’s look at some examples.

Using publically available analytics from Quantcast, I’ve collected audience loyalty stats for a number of publisher websites. Notice that native web publishers are generally near the top of the loyalty list, while print publications round out the button. The top traditionally print magazine website is The Atlantic, whose digital-fist approach is paying dividends both on and offline.

Reader loyalty by web property (February 2011 Quantcast #s)

Site % of visits from regular readers
Drudge Report 91%
CNBC 77%
Politico 74%
The Daily Beast 70%
Gawker 68% 67%
Atlantic Media Company 67%
The Economist 67%
Motley Fool 66%
Slate 65%
Tech Crunch 63%
New York Magazine 56% 55%
Newsweek 51%
Washington Times 50%
Fast Company 50%
Time Magazine 49%
Car and Driver 48%
National Geographic 45%
Kiplinger 44%
Christian Science Monitor 43%

Technology, the real Achilles’ heel for publishers, will put the final nail in the coffin. Analytics-driven ad targeting will allow advertisers to create their own walled garden audience segments within the open web—and at significantly more scale. For example, an advertiser could target only daily users of The Daily Beast who visit its homepage. Over at Rupert Murdoch’s iPad newspaper The Daily, the same advertiser might only reach a couple hundred people if recent subscription estimates are remotely accurate.

Publishers: wake up and smell the roses blooming outside your wilting walled gardens!

Does The Daily Beast have more grey hairs than

Last week, the two-year-old Daily Beast website and its editor Tina Brown assumed control of the faltering, near-octogenarian Newsweek. One of the first executive decisions of management was to put Newsweek’s website on the chopping block. “The Daily Beast is, arguably, the better, more vibrant web product. Their average reader is younger,” opined a blogger on, echoing the sentiments of The Daily Beast management.

While the Beast may be the better product, and is a younger animal in itself, my research shows its readers are older then’s. About half of The Daily Beast’s core audience is over 55, while just 27% of Newsweek’s are. This is according to rough demographic estimates made based on search keyword data complied by Yahoo Clues:

Age of Daily Beast vs. Newsweek Brand Keyword Searchers


This seems counterintuitive. Can you name one person who subscribes to Newsweek under the age of 50? But while it may be losing its paper and ink fans, Newsweek’s online brand appears to have a broader reach when it comes to age. This may be due to a strategic decision, years ago, to hook up with MSN/MSNBC. Through this special relationship, Newsweek content gets top billing on one of America’s most popular general-interest portals – extending its brand very broadly.

By contrast, The Daily Beast has build up an organic following online without reliance on massive brand recognition or portals. Rather than pushing its content out via the traditional, general interest media channels, it has pulled in a niche audience on its own. And these readers are, according to Yahoo’s numbers, older then Newsweek’s.

The other myth circling the media echo chamber concerns assumptions about Newsweek’s relative online popularity. True, it has more unique visitors than The Beast (7 vs. 4 million). And who outside of the NYC/Beltway chatter class knows about The Daily Beast? But the fact is search volume for The Daily Beast is about on par with Newsweek. Most of Newsweek’s traffic comes from MSN, and is not driven by brand-related search interest. Yes, more people do know about Newsweek. But advertisers care much more about brand engagement – and it appears the Beast is neck and neck with Newsweek when it comes to searches:

Brand-Related Search Volume


No doubt these numbers from Yahoo have inaccuracies (to learn more about them, see Yahoo Clues). But I’d wager they’re more on target than the intuitions of media critics who are often biased by the past, rather then what’s happening now on the internet. We like to build clear narratives, but when it comes to new media, there’s often no such thing. And when you do have a narrative, it gets torn down pretty quickly. RIP

Does contextual relevance increase user engagement?

Browse any major content sites today and you’re bound to find contextual relevance features whose aim is to keep you clicking. Read an article about global warming, for example, and you may be offered another link about green energy companies.

While such a concept sounds great in theory, does it really increase engagement? It’s difficult to make broad statements because every website’s goals are different. However, a recent experiment by the web team at suggests media companies may be overvaluing the utility of certain contextual relevance tactics.

In an A/B test of’s article recommendation feature box, the web analytics team discovered “…stories displayed from randomly selected categories outperformed the most popular stories and related stories.” In other words, users preferred content that spanned a range of disparate topics, appearing to debunk the theory that contextually relevant content increases engagement.

But Bloomberg’s analytics team did discover a winning technique in the testing process. The feature that out-performed the other methods was an advanced machine learning algorithm that analyses users’ past viewing behavior. This indicates there is an inherent complexity to user preferences that can’t be solved by more generalized solutions like related content or article popularity.

We are just starting scratch the surface of online psychology – and a lot of the assumptions we hold about user behavior will likely be viewed as archaic 10 years from now. The most successful web companies of the future will be open to questioning conventional wisdom with analytics and testing.

Facebook says we’re getting happier. Should we trust it?

whos1Every Who down in Whoville liked Facebook a lot

Over the past few years, an economic grinch has packed much of America’s wealth into his bags. But as stocks tumbled, layoffs mounted, and foreclosures turned neighborhoods dark, a remarkable thing happened: we became happier. That’s if you put any weight in Facebook’s Gross National Happiness Index.

Based on analysis of positive and negative sentiment in status updates, the GNH’s goal is to be “indicative of how we are collectively feeling,” says Facebook. The index’s ups and downs are often tied to major external events. When Michal Jackson died, the GNH plummeted. And every Christmas, GNH is off the charts. But stepping back from the daily rollercoaster that is America’s collective mood, we can see a long term trend of growing happiness.

facebook gross national happiness index GNH

While gradual, we have seen approximately a three point overall happiness increase over the past three years. And even more substantial are the year-to-year increases in happiness during Thanksgiving and Christmas, up about 10 points and 6 points respectively. What could be causing this swell in cheery status updates?

Some insights can be gained by segmenting positive and negative sentiment separately. As it turns out, there has been a more substantial drop in negative status updates than the increase in positive updates. This may indicate we are not necessarily getting happier, just less upset. Here are 5 possible reasons for the trend:

1. Facebook demographics are trending older

College-age students once dominated Facebook. All the stress that comes along with finals, finding a career and relationships can add to the negativity factor. Recently, older users have been joining in droves. These new networkers are more settled in life, simply delighted to be reconnecting with friends, sharing little details of their life, pictures of family, etc.

2. Bush went away

Right about the time of Obama’s victory, there was an upswing in the GNH. No more need to post vitriolic tirades against the war-criminal president. Facebook users had hope, and so did their status updates.

3. Americans hate to admit they’re unhappy

Maybe we’re more depressed than ever? As companies let go thousands and houses foreclosed, perhaps out-of-work Americans were too proud to admit anything was wrong? Indeed, when you have to pull yourself up by the boot straps, complaining is not the best place to start.

4. We’re evolving socially

Chronic complaining is unattractive, but some people didn’t realize it until they joined Facebook. The more people posted negative updates, the more they realized their friends simply didn’t care. In this sense, Facebook could be acting as a kind of positive behavior reinforcer, because that’s the trait that’s most rewarded in social circles.

5. It’s easier to complain when life is good

The things people gripe about on Facebook are usually petty – a long metro ride, a terrible TV episode, or a bad professor. If you’re like me, you post these types of updates from time to time, simply to vent. But when bigger life events arise, such updates can seem almost decadent. If anything, America is a little less spoiled than it was three years ago. That can’t be a bad thing.

What do you think? Are we really getting happier? Why or why not?

The Magazine Brand Crisis

Never again! That’s what most magazine publishers say to the rotten year that was 2009. With 25% average declines in ad pages, the industry simply cannot afford a repeat. But it’s not just the economy. As consumers flee print and flock to the web for content, it’s imperative that publishers get it right online.

While there are more online readers than ever, the fundamental problem for magazines is largely overlooked: magazine brand names are losing relevance — even on the internet.

Fewer people searching for magazine brands on the web

Of 140 magazines I analyzed using Google Insights for Search, only 31 publications saw increases in brand-name Google search volume in the last year. The average magazine experienced a 10% decrease in web search interest. And over five years (2004-2009), the keyword “magazine” has seen an astonishing 40% drop in its share of Google search volume.


With the proliferation of alternative content sources, the web eroding the brand name power magazines once held. People are no longer seeking out specific brands like Time, Newsweek or Discover. Instead, they use brand-blind information aggregators like Google. People may be reading more. But if readers don’t care where content comes from, publishers can’t claim the same value to advertisers on the web.

Here is a look at the change in Google search volume for some of the major magazine titles between 2004 and 2009:

What did People magazine do right?

What did People magazine do right?

Loyalty = brand advertising revenue

Print magazines are particularly appealing to brand advertisers because they reach a specific cohort of loyal subscribers every month. Kraft Foods, for example, might choose to place ads in Fine Cooking magazine for a year. They know their brand message will be repeated in front of a relatively constant audience 12 times. Not so on the internet.

When a Fine Cooking reader seeks out recipe information through Google, she might happen upon a random blog instead of Fine Cooking’s website. Kraft loses much of the control it has with the print publisher. But it does have another option: Instead of placing ads with a specific publisher, Kraft may chose to use an ad network across hundreds of websites and blogs. These ad networks can target specific demographic groups much like magazines. This makes Fine Cooking’s website much less valuable to Kraft.

Developing online loyalty

If traditional publishers want to survive online, they must develop new ways to their web readers loyal that are more effective than ad networks. This will give advertisers confidence they’re reaching a valuable audience that returns to view their ads each month – just like the print product. Some advertisers are even banning ad networks from their websites. But given the trends in brand name search volume, it seems publishers will be fighting an uphill battle.

A counter-argument – brands are dead

Some might argue that the internet has forever changed the way people interact with content. Brand loyalty as we once knew it will never again exist. Consumers have so much choice now, and it’s difficult to keep them around when cool new websites pop up every five minutes. So some publishers may decide to take a different strategy: forget about traditional brand advertising and go for volume-based direct response ads.

Direct response ads are different than brand advertising. They are about eliciting an action, as opposed to a soft and fuzzy impression that takes months to build in a consumer’s mind. Google has built its empire on direct response ads; they only get paid for the clicks. Imagine a magazine telling an advertiser it will only have to pay based on the number of calls it gets on the 1-800 number listed! It would never happen. Direct response advertisers care little about making you feel a certain way – they just want you to act.

Most traditional publishers are in no position to create a direct response-style strategy that can out-compete Google. Their strength is delivering the loyal audience that brand advertisers crave. But if you can’t keep a loyal audience, it’s the only option if you want to survive from advertising revenue on the web. Increasingly we will see publishers’ content strategies evolve to capitalize on the economic realities of the internet – and maybe some will beat Google at its game.

One thing is for sure: the magazine brand as we knew it will never be the same.

Can Rupert Murdoch Outsmart Google?


You may have recently read of media mogul Rupert Murdoch’s rumored plans to de-index his newspapers’ content from Google. “The Philistine phase of the digital age is almost over,” he proclaimed at the October World Media Summit in Beijing. He threatened that the likes of Google “will soon have to pay a price for the co-opting of our content.”

The casual observer might deem this idea mad, given that about 25 percent of The Wall Street Journal’s web traffic comes through Google. Indeed, comments on the popular social media blog, Mashable, lambast the 78-year-old industry titan:

“This guy is off his rocker.”
“What a greedy old fool.”
“Murdoch is just another in a long line of execs who don’t understand the technology.”
“Ok, this is probably the most ignorant thing I have ever heard.”


Sure, call the old billionaire foolish if you’d like. But there’s one thing Murdoch does understand: his newspapers may go bankrupt without a change in the online economic landscape. The business basics of the journalism industry today are grim:

• On the web, advertisers can now reach consumers very cheaply (Blogs, Facebook, etc)
• Traditional news publishers, who rely on ads to support costs, must compete at these cheap ad rates in the open market
• This revenue does not come close to covering the overhead of a major online news operation

In this business context, Rupert Murdoch can either work to protect his billion dollar empire, or let it go to the blogs, so to speak. Obviously he wants to do the former. So what’s his strategy after giving Google the finger?

A deal with Microsoft?

According to reports, Microsoft and News Corp have discussed a possible content deal on its fledgling search engine—errr Decision Engine— Bing. Desperate to increase its market share, some think Microsoft may be willing to pay for exclusive access to Mr. Murdoch’s content. But what about the lost Google traffic?

Analysts estimate that The Wall Street Journal makes about $15 million a year in ad revenue from Google’s search traffic. Given cash-rich Microsoft’s recent $100 million promotional campaign for Bing, it seems a no brainer to pay Mr. Murdoch well beyond a measly $15 million for an exclusive content opportunity.

Will the industry follow suit?

As one blogger put it, it’s a kind of prisoner’s dilemma. If one company leaves Google, the company that stays with Google will see its traffic and revenue increase. But if all news websites band together and jump ship, they could all collectively be better off.

Using elementary game theory, I’ve created a hypothetical example of how Mr. Murdoch might view the competitive business situation in favor of choosing a Microsoft deal. To simplify the marketplace, let’s imagine that there are only two players: News Corp (player A) and the New York Times Co (player B).

Currently, both companies allow their articles to be indexed on Google, generating in the neighborhood of $15M from that traffic. This scenario is represented by the upper left-hand box.

Now let’s say News Corp decided to move to Bing’s exclusive deal, while the NY Times Co remained indexed on Google. In this scenario (upper right-hand box), revenue increases for both. Why? Microsoft pays News Corp $40M per the signed deal. The NY Times gets $10M more in search traffic because News Corps content is gone from Google, giving it greater market share.

However, using logic, neither of the above scenarios would happen given the game’s conditions. Based on the revenue potentials, it makes most sense to chose Bing no matter what. Even though you’d make more money if you were the only one on Bing, it will not happen. Both are better off choosing Bing. While revenue will be slightly lower (their content is now competing for the same eyeballs on Bing, so Microsoft does not pay quite as much to each), it is still better than choosing Google.

Does this have any grounding in reality? Not really, because we don’t know what the conditions will be in real life. And according to an inside source familiar with the Microsoft negotiations, “the economics do not seem to be there for the common arrangement initially rumored to be under discussion.” One thing is for sure: the economics must change if News Corp wants to keep its newspapers. In an indication its willing to help publishers change their losing equation, Google today announced more options for news publishers to control their content.

Looking into the future

In America today, news publishing is a business. For these companies, the only right choice is the one that yields the greatest long-term profit. Given the rapidly changing media landscape, I’m not sure anyone can predict what will work best in 10 years. But I’m glad we have companies experimenting with different models. The sad part is that many ideas will fail, and more than likely, many journalists will lose their jobs in the coming years.

Old media paternalism is dead. Can we take care of ourselves?


[W]e have in this country two big television networks, NBC and CBS. We have two news magazines, Newsweek and Time. We have two wire services, AP and UPI. We have two pollsters, Gallup and Harris. We have two big newspapers–the Washington Post and the New York Times. They’re all so damned big they think they own the country.

Lyndon Johnson, 1969

If LBJ were still around today, he might be happy to know:

Yes, LBJ, things changed. The old media aristocracy has fallen. In it’s place: the internet and cable TV. Media consumption is no longer a three course meal served on schedule, but self-made snacks served up on-demand.

Apple has become the entree du jour, prefixing its all its media consumption products with an “i”– connoting this new self service reality. With the flick of a finger, one can easily filter out influences that might be too foreign, difficult to understand or just plain boring. This proliferation of personal choice can have the effect of stymieing our own intellectual development; we no longer need to listen to anyone but those who confirm our own psychological needs.

Old media has caught on to this. In response to the new consumer power, the besieged moguls are fighting back by giving consumers more and more of what they want: junk food for an insatiable ego. The rise of opinion-based programming on cable TV is evidence of this. Fox “News” has built itself around right-wing pundit personalities, while MSNBC has responded by taking a stake out on the left. Meanwhile, middle-of-the-road CNN saw its lowest ratings in decades.

When left to their own devises, people tend to consumer media that confirms and even amplifies their personal preferences (see my blog post about YouTube ratings). It’s becoming a much more personal experience. As further evidence of this new paradigm, Yahoo is currently running a $100 million ad campaign around the world that has one message: it’s all about you!


So the power of the old media has been transferred to us. In many ways this seems like a good thing – more voice to more people, the very basis of our democracy. But is this choice and freedom uncovering a strain in human nature we’d rather not face up to?

The old media aristocracy functioned a bit like a benevolent dictator. It used its monopolistic power to extol journalistic standards like truth, accuracy and trust. It was far from perfect, leaving many voices out. And thanks to new media, niche issues like gay rights are becoming increasingly recognized. But at the same time, so are problems like racism and xenophobia permeating the media. And in place of the paternal void, consumers are increasingly gravitating to rogue media voices.


I witnessed the tea party protests at the National Mall last month. Never have I seen a scarier political sight in my life. Egged on by Fox News pundit Glenn Beck, angry white marchers descended on Washington to spout ugly racist slogans. One older black man told me he had not felt so much hate since the 1960s.


No longer bonded to paternal media overlords, one might argue that we’re in a state of nature. This unprecedented freedom of information may be the ultimate test of the old Locke vs. Hobbes debate.