Category: Marketing Stats
July 26th, 2006

Anderson’s tall tail: critics wag the Long Tail

The sharks are circling the Long Tail on the web.

Lee Gomes in today’s Wall Street Journal throws a few spanners in the spokes of Wired editor Chris Anderson’s sensational Long Tail argument. 

Gomes took another look at the data Anderson uses to support his claim that thanks to the efficiencies of the Internet the sales of "misses"  will overtake the sale of "hits" over time.

– Anderson starts his book with an interview with digital jukebox Ecast. The CEO of ECast told Anderson that when ECast started 98% of the songs sold every month. Gomes checked in with ECast and discovered that since Ecast’s inventory has increased, the quarterly no-play rate has jumped to 12%. Rhapsody, an online music business similar to iTunes and Ecast, has a 22% no play rate. 

– Although Netflix and Amazon do make the sales and distribution of misses much easier, it could take decades for the misses to outsell hits. Incredibly, according to the Journal, Anderson "told me that he didn’t have any examples of this actually occurring." Ouch.

The Journal’s analysis of Amazon sales notes that only 25% of sales are in the tail; while incredibly 75% of revenue comes from 2.7% of sales. (On a separate note Amazon reported a 58% profit drop this week — um. Not good. Investors expects stellar earnings from the Seattle bookseller after so many years in the red).

– The Journal also reports that iTunes sales shadow Billboard hits. This informed source adds: " Its a hits business. The data tend to refute the Long Tail". Read Gomes’ full article here: "It may be a long time before the long tail is wagging the web".

Tim Wu, a Law professor at Columbia, also took Anderson to task in Slate yesterday "…like many business books, The Long Tail commits the sin of overreaching"… he continues:

"What are the Long Tail’s limits? As a business model, it matters most 1) where the price of carrying additional inventory approaches zero and 2) where consumers have strong and heterogeneous preferences. When these two conditions are satisfied, a company can radically enlarge its inventory and make money raking in the niche demand. This is the lifeblood of a handful of products and companies, Apple’s iTunes, Netflix, and Google among them, all of which are basically in the business of aggregating content…."

"But it’s important to remember that many industries don’t rely on the weird economics of information products. Take the oil industry, which Anderson doesn’t discuss, but whose significance is obvious—compare Exxon’s $371 billion in revenues in 2005 to Google’s $6.1 billion. The Long Tail doesn’t seem to tell us much about the future of the oil biz. It’s not really clear how Exxon might benefit from expanding the types of gas it makes available at its service stations. It would cost Exxon a lot to install extra pumps, and few people have well-developed tastes for types of gasoline. Since it’s not easy for Exxon to reduce its inventory costs, product diversification is expensive. There might be long lines in gasoline retail, but there’s no Long Tail."

"The Long Tail also sometimes doesn’t work in its home category: the information-technology industries. The key issue is the question of standardization. Sometimes consumers want a diverse set of product offerings. But sometimes they prefer a standard or compatible product. Most of Anderson’s examples are content firms, where product diversity is almost always a good thing. But in the information-transport industry, standardization is usually more important. Do people want 10 different types of (incompatible) Internet connections? Or just the fastest one they can get? How about 30 types of (incompatible) Ethernet cables?"

Read the full article here: The Wrong Tail: How to turn a powerful idea into a dubious theory of everything.

Anderson has given us a new lens to consider Internet-powered commerce. His theory may not quite rise to the heights demanded by professional economists, but its hard to deny that he is on to something.

Just witness the strength of the debate.

We’re still in the early days of the web: it takes bravery and chutzpah to stick your neck out and predict where its going. I’d watch this Wired editor if I were you.

Besides, have you ever seen a business trend birth more bad puns?  Wu’s closing comment is a classic:

"The Long Tail isn’t useful as a theory of everything. It is best and strongest when it helps us understand what’s happening to our culture. It shows, graphically, the difference between the mass culture we’ve had, and the folk culture we’re bringing back. That’s an achievement worth celebrating, and it’s why the Long Tail can leave us feeling like cavemen looking at a map of the world for the first time. But the book should come with a warning: There’s more to this economy than chasing tail." Booyakasha.

Note:  You can catch Long Tail author and Wired Ed-inChief Chris Anderson at our client, Golden Gate University, on Weds. September 20. Click here for the details:

July 22nd, 2006

TV text messaging: show me the money and I’ll show you a trend.

American Idol was much more than a cultural phenonmeon. It was also the first smash hit for text message voting and, it appears, a major weapon for the networks in their battle to keep viewers from watching programs on-demand, via DVRs like Tivo.

"Idol" generated over $6 million in texting fees last season, from 64 million text messages (each cost $0.10) up from 7.5 million messages in 2003.  Broadcasters typically get 40% of show text revenue, carriers — Cingular, Sprint and Verizon — keep the balance.

Moved by the found money in texting, and agreements to fix toll charges as high as $0.99, sufficient cellphone network bandwidth and widespread carrier participation, more shows will seek viewer input next season reports Li Yuan in July 10, Wall Street Journal "Televisions New Joy of Texting."

"Big Brother," "Deal or No Deal," "The Country Music Awards," "America’s Got Talent" and the upcoming "Hell’s Kitchen" will all drive the new wave of consumer participation. Dozens more shows are expected to follow suit.

And that pleases advertisers and the networks for a number of reasons:

Greater interactivity gives broadcasters a weapon to battle the rising tide of DVRs, because it ensures more viewers watch live shows rather than recorded versions. Providing a bigger captive audience for advertisers (and another metric to gauge viewership.)
Overall viewership jumps as more people stay tuned to watch the show from beginning to end.
Voting drives loyalty: CBS notes that some voters joined Big Brother’s online fan club, adding to network online revenue and ensuring more return to view the show every week.
Text voting helps spread the idea of texting, further strengthening carrier revenues.
And texting strengthens the bond between viewer and show. During last season’s "Idol" finale, my family texted continually — and the interactvity spurred conversation throughout the week as we waited for the results of the run off between the spasmodic Tyler Hicks and the glamorous Catherine McPhee. 

If you want to know where the Web is headed, just follow the money. 

At a buck a message, not only can we can expect to see more shows make room for texting in the future, we’ll also see more radical format changes in shows as producers work to maximize text revenue.

Just as banner advertising opportunities multiplied on web pages as publishers made the most of selling the same eyeballs, and news broadcasters agreed to clutter screens with the news ticker, we will see TV producers push interactive voting to the limit of effectiveness.

Similarly we can expect dramatic changes online as web publishers and carriers rush to grab incremental text revenue from less revenue-rich sources like email, VOIP and IM.

Remember where you read it first.

July 13th, 2006

Web makes a monkey of the Networks. Virals ads twice as effective as Superbowl advertising.

Picture_11_3 The Wall Street Journal reports that Careerbuilder’s recent Monk-e-mail email program has been played 44 million times and attracted 6.5 million unique visitors to the site. As the budget was "around" $200,000 it compares very favorably to Careerbuilder’s ads on the Superbowl.

In contrast, the TV cost around $5 million to air, reaching a maximum of 90 million viewers.

Mortar’s advanced mathematicians figure each Superbowl viewer cost 5 cents. A view of the Monkey’s provided a web view for half that (2.5 cents).

And unlike the Superbowl ad Monk-e-mail has a much longer tail — its still pulling visitors and generating word of mouth a year later .

No wonder the venture capital community is wondering why Madison Avenue is taking so long to move its dollars from broadcast to the Web.

July 6th, 2006

Most web advertisers hit with online fraud. New study marks bogus clicks as $1.3bn problem.

The latest estimate for click fraud indicates taht more than $1.3 billion of keyword advertising sales are fake, according to a study released yesterday from researchers Outsell (the full study is $495).

Highlights (as reported by SFgate July 5, 2006):

–  14.6% of clicks are fake.
– 75% of advertisers have been impacted at least once (One wonders if those who say they haven’t probably are simply unaware of it).
– Nearly a third of advertisers have reduced or stopped spending on click-based advertising. An additional 10% intend to curtail spending.

"Outsell’s survey was based on the responses of 407 online advertisers representing a cross-section of U.S. business. Their spending ranged from several thousand dollars online annually to more than $10 million". SFgate.

Outsell cleverly heralds their reports as the beginning of the "Don’t ask, don’t tell" era in keyword advertising.  Its clear this issue is going to be a major headache for the search engines.

Mortablog readers are reminded of our post "BlackFoot calls time out on click fraud."

The opinions expressed in MortaBlog are not necessarily those of the author or anyone else at the Mortar
for that matter.  Just who owns them is kind of unclear really.  If you
do find someone who will own up to them for sure, let us know.


June 29th, 2006

Are Microsoft adCenter conversions really 57% better than Google?

In ads out this week, Microsoft claims conversions from its  new adCenter search advertising system are 57% higher than Google.

MSN pulled the stat from this January press release from online analytics company WebSideStory. Further review of the release reveals:

  • For the month of January, AOL Search generated the best conversion rate
    at business-to consumer e-commerce sites (6.17 %), followed by
    MSN (6.03 %), Yahoo (4.07 %) and Google (3.83 %).
  • The study includes traffic from
    both organic and paid keywords.

Frequent readers of Mortablog will have noted that we criticized Google last week (See The Google Touch).

So, in the interests of evenhandness, let’s review the merits of Microsoft’s pitch for your search business:

MSN is talking  ratios not absolute numbers. MSN is not measuring the number or propensity of visitor’s clicking through to a site from organic (so-called natural) or paid links: they are calling our attention to visitors likelihood to buy once they land on the site. MSN claims that visitors from paid listings are more likely to buy if they came from MSN than if they originated at Google.

AOL Search took top honors, not MSN. What are we to make of the fact that despite MSN’s proud boast, WebSideStory gives the conversion crown to AOL Search and not Microsoft? Unsurprisingly, MSN is positioning itself against market leaders Google (who incidentally have the lowest conversion rate) and Yahoo! (the second worst converter). Savvy advertisers will remember that AOL Search draws its listings from Google and AOL itself. One wonders if we’ll see a riposte from AOL Time Warner?

Much has changed at MSN since January
. WebSide’s research was conducted in January 2006 before the official launch of adCenter in May and subsequent recent changes to MSN Search. Back then, Yahoo! was still providing some of the paid search listings on MSN Search. The study also predates MSN’s planned summer test of contextual advertising (see here). These developments have the potential to impact the reaction of visitors to listings.

Mixing conversion rates from Paid and Organic search does my head in. Look, advertisers can’t purchase a Page One rank in Google or Yahoo! or MSN for that matter. Organic links are organized and displayed according to algorithm and bots. Placement is not influenced by ad spending. Armies of SEO consultants optimize organic search by manipulating page code and encouraging visits and links from other sites. 

On the other hand, Paid listings are actively managed by automated advertising systems on each search engine and by advertisers themselves. Poorly performing ads are dynamically pruned and adjusted; top performers are pushed to the limits of effectiveness.

ComScore studied conversions across paid and organic keywords in April 2004 and concluded that paid search was ten times more effective at driving conversions than organic. (Read more here).  But, then again, advertisers pay for each click on paid search–whereas organic clicks are free.

Consider that for the VAST MAJORITY of webmasters organic traffic will exceed paid by a significant margin. The bottom line: sites with ten times more organic traffic than paid will register the same number of sales even if the conversion percentage is lower: rendering the whole issue of conversion percentage kind of moot.

MSN searchers like to shop. A better study for MSN might have been iProspect’s May 2004 review of search engine traffic. It indicates that MSN visitors are more likely to click on paid links than organic listings. The broader implications of iProspects’s work is that MSN is more Shopping than Search–and that would harm Microsoft’s long term vision for MSN as a contender to Google. (It may also say a little something about how much users trust products from Redmond). 

The WebSideStory participants were using HITBOX optimization technology (from WebSideStory): The WebSideStory release continues "One other important consideration to note in this study… is that the conversion rates are likely higher than
industry averages because the sample sites are using best-of-class web
analytics to improve their search engine marketing and optimization
". So, advertisers have to use WebSideStory’s HitBox system to get these numbers.

Microsoft’s ad certainly caught my eye. But it hardly stands up to serious scrutiny. Let’s hope that adCenter is not similarly flawed.


The opinions expressed in MortaBlog are not necessarily those of the author or anyone else at the Mortar
for that matter.  Just who owns them is kind of unclear really.  If you
do find someone who will own up to them for sure, let us know.


Update: Paul A. sent me this email:

Mark.  Sites with 10x more organic than paid ?   Sure.  There may be one or two in the universe.

But that’s not an argument against paying when you are not happy with your organic traffic.  In fact, why are you saying that conversion rate is a moot point.

I would gladly pay up to x % of sale for the sale itself.  ( x% being my acceptable cost-of-sale through digital marketing–different for every product and markeing plan).  The only way that x% becomes 0 –which would make it moot –is if I am already at my production capacity through organic, non stimulated sales.

Remember, conversion % relates to gross payout for clicks divided by gross sales.  It definitely matters and it matters most to the marketers with low organic traffic.

I think I have to concede this point to Paul. Paid conversion rates certainly matter to Webmasters of small or unindexed sites who received low levels of organic traffic — Mortarmark.