broadstuff

You would have thought that after Harvard, McKinsey, Google, Amazon, Will Page and Apple debunked it, in fact even after the original author climbed down, the Long Tail Theory would be dead by now*. But no, Techdirt rides in to its rescuscitation:

There have been a series of criticisms to Chris Anderson's concept of "The Long Tail" lately. While most don't hold up under scrutiny, a few have made some good points that don't actually go against the long tail concept, but may adjust some of how people understand it. Of course, people love to jump and declare something completely dead, so there have been a few stories making the rounds claiming that the long tail era has been disproved.

Not hold up to scrutiny cough! splutter!. Yup, cancel your subscriptions to the McKinsey Quarterly and Harvard Business Review and buy TechDirt Anyway, the new argument is here:

eMusic has published a report noting that, indeed, their own sales show a pretty clear long tail distribution, with at least 75% of all of the music they stocked being sold at least once in 2008. Once again, though, it comes down to the filters. eMusic has some good features (they could be better, honestly) to help people find new and obscure music -- and that helps spread interest to new acts. So, once again, it appears that the long tail is still very much alive, but it does still depend on the filters being used.

Those last 3 sentences are actually by and large true - the Long Tail is actually alive, and it depends largely on its filters. But is is not very big, and never has been, and that is essentially the claim the economists debunked from the original Long Tail book, which held that it was a much larger opportunity than had previously been supposed.

From Pareto's 80/20 law (and long before in fact, the Phoenician traders of yore knew of the principle), people have long known that consumption of nearly everything is a power law, with the bulk (the 80%) of the consumption going to the top sellers (the 20%) of the goods. In fact, in hit based businesses like Media, it can be much worse - a 95 / 5 split is not uncommon.

In fact, the old inventory theory used to define the "A" class (the hits), the "B" class (the movers) and "C" class - the Long Tail. However, e-Music above defines "long tail" as anything not in the Top 40, which is like arguing that anything not in the top few % of sellers is "Long Tail" - ie the Long Tail begins at the neck .

What is true is that if one can make the back-catalogue cheap to store, easily searchable, and cheap to distribute, it makes it worth the while of retailers to eke that last tail-off few % of revenue from the market. And digital music media is 2 orders of magnitude cheaper, so it is worth their while.

But this is not a revolutionary new business model, unless you can build a system to become a bottom fisher across multiple long tails, like eBay and Amazon did. To make money you need serious scale, and low transaction costs. Its life as we've always known it.

* To be fair to Chris Anderson, his book was a well argued hypothesis that was over time shown to be not always correct, he is far from the first (or last) business theorist to whom this has happened to - the ongoing momentum of the Long tail is more from acolytes who are still trying to maintain commercial practices from it.

One of the big shifts in value in the music world was when Napster, and almost all Web services that followed, broke the music "bulk package" - the album - into its parts (songs) and distributed them separately, letting them find their own value. Work we have done in the Future of Online Video implies that a similar practuce has to occur for it to take off (and to an extent it has, via YouTube but search and legality are both dubious) and be usable. The question is, who will do it andhow will it be paid for?

Thus this post from the NYT about Discovery Chanenls re-purposing of its media into clips is interesting:

Discovery owns the global rights to the content it broadcasts. Because the company owns 13 networks in the United States, including TLC, Animal Planet and the Science Channel, it has untold thousands of hours of footage. And because many of its TV episodes are timeless, the clips can still be relevant to Internet users years after their original broadcast.

“Cheetahs are still killing gazelles the same way they did 3,000 years ago,” Doug Craig, the senior vice president for digital media production, said, “and on top of that, we don’t have to pay residuals.”

To that end, Discovery has added six temporary employees to “maximize the library,” Mr. Craig said. They are repackaging old shows into short clips for the how-to Web site HowStuffWorks, a recent acquisition by Discovery, and Discovery’s other platforms.

Two important points here - timeless content retains its value over a longer sales cycle, and it is emerging that "inform" and "educate" are key usages of online short form media, as opposed to the TV mantra of "entertain". Also, Clips and Ads go together even in tough times:

While the online advertising market has slowed considerably because of the weak economy, the executive vice president for digital media at Discovery, Josh Freeman, said advertisers’ demand for video remained robust. Asked whether the ads were lagging behind the content creation, he responded: “In video? Not a chance.”

Next steps are better metadata and search, and we would postulate the emergence of trusted New Media channels to watch (and take Ads) rather than the cr*pshoot that is YouTube in its current format.

Silicon Alley Investor asks why Google Knol hasn't been culled in the current Googlekrieg on bits of the business that burn cash:

This weekend, Google announced a Knol contest with $1,000 for the best entry and some cross-promotion with dummies.com.

We're not sure what the thinking here is. We're still not even sure what Knol is for. For a how-to, like performing car maintenance, we'll go to a site that's run by car-fanatics. For general knowledge, Knol can't come close to touching Wikipedia. If anything, Knol seems to have a reputation for nothing so much as plagiarism, ignorance, and self-promotion.

We admire Google's willingness to experiment with new ways to build cool (and potentially profitable) features onto their existing service. And we admire Google's willingness to realize when their experiments have failed and shut them down. So why still back Knol?

This brought a rapid response from Google's Matt Cutts, arguing that it is still alive and kicking. He notes 4 things we need to know about Knol, as possible reasons for sparing its life:

- Google Knol does not receive any sort of boost or advantage in Google’s rankings
- The conventional wisdom changed from “Google will give Knol unfair boosts in ranking; it will dominate the space!” to “Oh, Knol gets so little traffic that it’s not a success.
- The Knol team is not standing still.
- Knol is that when you want to write a quick article or put some information on the web, Knol is a great place to do it

Not exactly compelling reasons for continuing life - a biz model shift from wiki-paid-here to deli.cash.us.

When it was launched, we couldn't see how Knol would work in its current format (see analysis here) but I suspect that it hasn't been around long enough yet to be on the chopping block, would guess its probably in a set of projects that have been given X (say 6) more months to prove their viability.

A quick search on Knol brought up a direct-ish entry for Turkey, but none for Guillotine

HipMojo on the valuation of Twitter at c $250m via a 3rd funding round of c $20m:

Twitter is making the same mistakes Facebook did, which are:

- raising money instead of generating any,
- letting the valuation get ahead of realistic business prospects which will make any M&A nearly impossible,
- thinking that the “success” its had with its obsessive, early-stage adopters can be mirrored to the broader world,
- will obsessively create valuations based on online advertising estimates even though the company is at best a communications and/or e-commerce play,
- mistaking a API for others to use as a business model
- encouraging other businesses to build applications on top of their grid, without itself knowing that their business is/will be.

There are some differences:

- Twitter is more of a mobile app than Facebook is, and mobile, while wildly speculative, is all about hype and fads with very little to show for it. On top of this, the wireless world remains a largely closed one where to become a true success you need to operate in the world of carriers, something that is impossible to pull off.

- If Facebook really wanted to, it could put Twitter out of business by pushing their own status update feature (more people have a Facebook account than Twitter will ever have).

- Lastly, while over time a majority of Web surfers might have a Facebook profile and could stay somewhat engaged with it to varying degrees, the same simply cannot be said about Twitter.

I'm normally a fan of Ashkan Karbafrooshan's stuff (he writes HipMojo) but this time I must disagree in parts - Twitter is not a social network per se, its more of a comms application system (think email 2.0) in my view, and more real time - that puts it in a different bracket to Facebook. As to raising money, they need to do it now to give them the breathing space to find a business model and there are signs that it is mainstreaming as TV personalities and Slebs come on board. Granted its risky, but the funding they are taking is pretty small beer for something that may be the next Email - if anything, this gives me some faith the "venture" bit of the VC market is still working.

As to the point about mobile - personally, I think thats how they should try to make money - people actually pay for mobile, putting a small % of the traffic it generates going into Twitter's pocket is very scalable.

Interesting essay on the future of books in the New York Times today, showing that the outcome of the recent Google settlement is far from clear and that the issue of copyright has veen wrested from the ideals of protecting knowledge to one of protecting profit. The point that struck me was the 400% inflation in the time copyright applies over the last 300 years:

...copyright was created in Great Britain by the Statute of Anne in 1710 for the purpose of curbing the monopolistic practices of the London Stationers' Company and also, as its title proclaimed, "for the encouragement of learning." At that time, Parliament set the length of copyright at fourteen years, renewable only once. The Stationers attempted to defend their monopoly of publishing and the book trade by arguing for perpetual copyright in a long series of court cases. But they lost in the definitive ruling of Donaldson v. Becket in 1774.

US law by and large backed up the British concept of 28 years, but in the last few decades corporate influence has swayed legislators considerably:

According to the Sonny Bono Copyright Term Extension Act of 1998 (also known as "the Mickey Mouse Protection Act," because Mickey was about to fall into the public domain), it lasts as long as the life of the author plus seventy years. In practice, that normally would mean more than a century.

So - extension for a maximum of 28 years to c 140 years - this is hardly what one would call progress over 300 years. The current agreement with Google, while it stops them pwning the entire world's literary output, still essentially just extends the cabal of those who do have the rights by one extra big corporate. And no ordinary corporate - as the article notes:

No one can predict what will happen. We can only read the terms of the settlement and guess about the future. If Google makes available, at a reasonable price, the combined holdings of all the major US libraries, who would not applaud? Would we not prefer a world in which this immense corpus of digitized books is accessible, even at a high price, to one in which it did not exist?

Perhaps, but the settlement creates a fundamental change in the digital world by consolidating power in the hands of one company. Apart from Wikipedia, Google already controls the means of access to information online for most Americans, whether they want to find out about people, goods, places, or almost anything. In addition to the original "Big Google," we have Google Earth, Google Maps, Google Images, Google Labs, Google Finance, Google Arts, Google Food, Google Sports, Google Health, Google Checkout, Google Alerts, and many more Google enterprises on the way. Now Google Book Search promises to create the largest library and the largest book business that have ever existed.

So, who will raise the beacon of anti-trust then?

Twitter has been pimped big time on the Return of Jonathan Ross show, a c 3 day extravaganza of PR hype culminating in his first show after being booted off air for being a bit of a prat Anyway, Twitter was one of his chosen vehicles for personal restitution, and the faithful fans have been hanging on every text as the buildup to the show approached its zenith - for example, as the BBC noted:

Fry, posting on Twitter, the micro-blogging service, said: "JR and I discussed Twitter. Hope it makes the cut. Cruise charming."

Ross used his own Twitter account later, writing: "Hello. That was a fun day and nice to be back. All the guests were so lovely, and funny. Twitter discussed - Tom [Cruise] might sign up!"

And here they are talkin' Twitter on TV.

Anyway, that paled into insignificance compared to the twittergasm actually during the show. The BBC notes that:

...one member of the audience was asked to leave after they were suspected of recording the show.

They have no idea - just read this for a view of The Twittergasm (people with delicate stomachs should look away, there is a dangerous level of psycophancy).

it is illustrative of the mainstreaming of Twitter in the UK that Stephen Fry has more twitter followers than US hyperhypecommunicator Robert Scoble (60,600 vs 50,44). Here, as they say, comes everybody - and there, as they say, goes the neighbourhood .

(One of the discussions at Amplified 09 was whether the current users of Twitter will stand the flood of newbies coming onboard as it mainstreams following celebrity endorsements, or will they go somewhere else)

Here's the roster (data from Comscore):

Top 15 countries, by Internet population:

1. China: 179.7 million
2. United States: 163.3 million
3. Japan: 60.0 million
4. Germany: 37.0 million
5. United Kingdom: 36.7 million
6. France: 34.0 million
7. India: 32.1 million
8. Russia: 29.0 million
9. Brazil: 27.7 million
10. South Korea: 27.3 million
11. Canada: 21.8 million
12. Italy: 20.8 million
13. Spain: 17.9 million
14. Mexico: 12.5 million
15. Netherlands: 11.8 million

Worldwide Internet Audience

Asia Pacific: 416 million (41.3%)
Europe: 283 million (28.0%)
North America: 185 million (18.4%)
Latin America: 75 million (7.4%)
* Middle East & Africa: 49 million (4.8%)

Thats about 20% of world population. China and India have tiny penetrations as % of population, guess where the major growth will be from now on..............

Just in case there was any lingering doubt that running a business off free cloud services today is a cr*p idea, Google rammed the message home - TechCrunch:

Google says the vast majority of the 1 million businesses that use Google Apps opt for the free advertising supported version. To make the free option less attractive they’ve been quietly lowering the number of user accounts that can be associated with a free account. Now as businesses grow, they’ll be forced to move to the paid version much more quickly than before.

.....

When Google Apps first launched in August 2006 it was free and described as “a service available at no cost to organizations of all shapes and sizes.”

Free for everyone lasted until February 2007, when Google announced a premier edition of the service with more storage and an uptime guarantee. The cost was (and is) $50 per user per year.

When Google Apps first launched up to 200 user accounts could be created for each business under the free version. But that limit was quietly reduced to just 100 user accounts. And then when the reseller program was announced earlier this month, the limit was cut in half again, to just 50 accounts.

Things, as they say, can only get better - not. The price of losing control is eternal vigilance. The good news is for competitors, as commentor Ian Hendry pointed out:

Business application ISPs worldwide will be breathing a sigh of relief. This sets a precendent for paying for business apps even if delivered online as SaaS.

I share the view that ads while trying to get business done using apps that just happen to be delivered online are a nuisance. People will pay if the apps are important — Salesforce.com has proven that. If anything, I expect Google Apps to appear more credible because they DO have a fee paying model.

Another freeconomic bubblet bursts......

Chart shows excess wages over mean for financial sector from NYU- (hat tip Paul Kedrosky) as a predictor of recessions / depression it...well, the implications are clear.

Though what I don't understand yet is exactly how the conditions arose for the rapid climb in the last 15 years. When I was newly out of college, other professional salaries were roughly on a par with bankers - but the difference skyrocketed from c 1990 on, leading to probably one of the biggest misallocations of talent and resource in the West in the last 200 years (ie ever since the Age of Revolutions took hold and Ancien Regimes were toppled or reformed) and I would hypothesize has contributed hugely to the US's (and UK's) loss of competitiveness across many sectors in that time.

As to why, once it starts the system dynamics are pretty unsubtle - as the Economist noted today, the main problem is that banking also suffers from adverse selection issues as well, as fear moves to greed in the boom time:

...as a boom takes its course, fear is supplanted in what a senior quant at an American bank calls the “Cassandra effect”. The more you warn your colleagues about the tail risks—the rare but devastating events that can bring the bank down—the more they roll their eyes, give a yawn and change the subject. This eventually leads to self-censorship. “The system”, he says, “filters out the thoughtful and replaces them with the faithful.”

And it was an escalator....

.....it has seemed as if everyone in finance has wanted to be someone else. Hedge funds and private equity wanted to be as cool as a dotcom. Goldman Sachs wanted to be as smart as a hedge fund. The other investment banks wanted to be as profitable as Goldman Sachs. America’s retail banks wanted to be as cutting-edge as investment banks. And European banks wanted to be as aggressive as American banks. They all ended up wishing they could be back precisely where they started.

Greed without Governance, allowed to build over N years, equals depression, and its absolutely predictable - now isn't that depressing....

There has been an interesting resurgence of interest in community allocated "social capital" systems, driven by the rise of social networking, a concept sometimes known as "whuffie" from a Cory Doctorow novel, Down and Out in the Magic Kingdom, imagining a world of a post shortage economy of plenty - Wikipedia on Whuffie:

The usual economic incentives have disappeared from the book's world. Whuffie has replaced money, providing a motivation for people to do useful and creative things. A person's Whuffie is a general measurement of his or her overall reputation, and Whuffie is lost and gained according to a person's favorable or unfavorable actions. The question is, who determines which actions are favorable or unfavorable? In Down and Out, the answer is public opinion. Rudely pushing past someone on the sidewalk will definitely lose you points from them (and possibly bystanders who saw you), while composing a much-loved symphony will earn you Whuffie from everyone who enjoyed it.

Now I've read quite a lot of the theory behind this, and it all sounds marvellously idyllic - but I am increasingly coming to the view that reputation systems driven by unmediated opinion are very dangerous (how does one ensure the wisdom of crowds is not actually the madness of crowds). The reason for this view is my experience with our karma voting system on this blog, which I've now removed. The only real lesson to take from analysis of our karma voting system over 2 years is that:

(i) Votes are usually a tiny % of views - 5% would be high.

(ii) There is one thing that drives voting far, far more than anything else - popularity at the time, ie adherence to the zeitgeist (another word for mob opinion)

And, as is palpably clear watching behaviour on social networks like Twitter etc, people understand this "dark side" only too well. The chart below explains the game theory of what really happens in "whuffie" type reputation systems:





Now, in theory, the whuffie system is unbiased and awards social capital - karma points - to people who incline to the right of the table, do the right things, speak truth unto power, are genuine etc etc. The problem occurs where the zeitgeist - the agreement about what is good and proper - can be shifted by the beliefs of the people within the system. This is the reality of the human condition, there are always trends, beliefs etc that are not true (often promoted for self interest), and the opportunity to play games exists.

I call these game players the "karma chameleons" - people who change their views with the zeitgeist to gain (game) the collection of karma in the system - to always espouse the popular view is a very useful strategy. This is what we saw with our karma system on the blog. Write a post on something that is the popular view, I'd get high karma votes - no matter if it was correct or not Contradict the popular view, and negative votes pour in.

The karma chameleons gain in the game at the expense of those who are espousing views that are not part of the zeitgeist. The risk is where the zeitgeist view conflicts with an unpopular but true view - the people who espouse this latter become "prophets" (ie they will only be respected in other virtual countries with unbiased value system). The problem with this position is that there is no profit in it, even if it is correct for the system's health - so those who would espouse the truth - and still profit - move to the "truth economist" position - ie to gain whuffie but still speak truth, you have to be "economical with the truth" and just not enter into large areas of discussion. One can think of a number of areas of discourse today where these areas exist, to the high risk of many in some cases.

Now bear in mind, in the social network capital system as imagined, whuffie is all - it defines your wealth, which defines status, access to resources etc within the ecosystem - so this game is played hard and to win.

And this is the risk with these systems, if they become the only way in which people accrue status in a system and all rewards are allocated by crowd opinion. This is even more exacerbated in social media by the power law dynamics it works under - as was noted the other day in a humorous (but sharply true) rant about Twitter "Social Media Experts", the Big Dogs in the system tend to attract acolytes who "clamp on" to any behaviour that is espoused by the existing high whuffie individuals (game being played - karma chameleons become acolytes of the highly placed to gain whuffie at expense of weaker players).

This ecosystem thus, without some form of strong mediation, resembles not so much the democratic system that the whuffophiles imagine, but is potentially more like a Feudal system with robber barons and their bands roaming round the social net mugging defenceless digital peasants (or at the very least lording it over them and forcing acquiescence to their ways).

How must the mediation work then? In the real world it works because Real World Whuffie comes from a variety of sources, the social mob cannot for example vote my salary this month to zero if I say things they don't like on my blog, because my salary is paid via an alternate form of capital. Also, transaction costs are high, so real life society - in the majority of cases - doesn't prosecute or throw people into prison lightly, and even then usually there is some mediation via a legal system that all are aware of. In fact, as Robert Wright points out in his book Non Zero, human civilisation has been a long process of trying to reduce the impact of just these sort of big man/mob rule based systems. This is not the case in today's virtual social media, which resemble more (in architecture) the closed kingdoms of despotic potentates (and capricious ones in some cases) - if you flee them, you leave all your accrued digital wealth behind. They are just too new, we haven't yet had the time to develop the tools to civilise these virtual environments yet.

OK, OK - I exaggerated to make the point, I'm coming at it from the point of view of imagining what it would be like if a whuffie based system was the only game in town, and you couldn't leave it, which is not the case today. But you can see the behaviours I have described in any social media system today - take the Twitter lynching of journalist Sarah Lacy last year for example ..

Anyway, thats the long winded reason that, from today onwards, we won't be using karma voting in future