Despite the arrival of fibre connections to the outside world in recent months, satellite is still the most cost-effective means of getting ISP traffic across the border. This is mostly because satellite bandwidth can be purchased in simplex, while buying fibre bandwidth involves paying for a massively under-utilised outbound connection. Obviously the satellite connection boosts the latency, so there’s a lot of trouble doing QoS so that customers who need the latency guarantees get routed to fibre.

The fact that satellite bandwidth is competetive at all should immediately tell you that bandwidth is a significant cost to the business. The economics of selling to a market of relatively poor people don’t stack if you have to buy relatively expensive bandwidth to serve them at a ratio of anything like 1:1. Caching and mirroring would seem to help, but it isn’t 1997 any more: most of what people do on the web isn’t static. You can’t cache Facebook. Even seemingly static pages often have enough dynamic content that you can’t reliably cache them. Mirroring the likes of Google search and YouTube might help, but you can’t get far without negotiating with the big boys, and they may not have time for such a small market.

Traffic shaping is another major opportunity to save bandwidth. No matter how unpopular the idea might be with Californian Slashdotters, in these circumstances the only alternative to an ISP that attempts to prioritise email over bittorrent at peak times is an ISP that goes bust and provides no service at all.

In the never-ending quest to get page views, the choices writers and editors are making to attract eyeballs and drive traffic are creating a new breed of low-brow, gimmicky disposable content.  At its best it adds little insight and at its worst amounts to a slimy bait-and-switch (catchy headline, nothing to say in the article).

It’s the new clutter. The article itself has devolved into a flashing, animated pile of fluff. The casualty of the rat race towards ad impressions isn’t just crappy layout and thoughtless art direction. It’s awful and useless content.

To me, this seems to be an all but inevitable result of attempting to fund the majority of the web publishing economy with advertising, which in turn is a natural outcome of the post-2000-goldrush exuberance that in the future, everything will be FREE.

The problem with advertising is that a casual or dissatisfied eyeball pays off just as well as an engaged, satisfied eyeball—quite possibly the former pays better, since a dissatisfied mind is more likely to wander away from the content looking for gewgaws to purchase. People (and companies) respond to this incentive, and attempt to gain maximal payoff with minimal effort.

Compare the open web with the content in Amazon’s e-book store (or any paid-content service): sure, there are plenty of bad books, but the vast majority are honest attempts to provide something of value, even if the author’s skill is found wanting.

Curiously enough, it’s not the lack of income that causes this: in places where advertising isn’t possible (e.g. wikipedia, feedbooks etc.) the average quality of content is higher. What seems to be important is not the size of the incentive, but the direction in which it pushes.

Currently some of the discussion in publishing centers on the fact that music is now sold in $1 downloads rather than $15 albums. But this too is an accident of circumstance. The length of a CD album is not a golden standard of the quantity in which people like to enjoy music, but one person’s personal taste: Sony vice-president Norio Ohga wanted a CD to hold the entire of Beethoven’s 9th symphony, and so for over a decade that is how we purchased our music.

Current commentary sometimes seems to imply that the music industry allowed the business model to be shifted to single-track purchases (to its cost) due to poor negotiation, and that the print industry might avoid making the same mistake. This is a delusion: certain changes in technology simply invalidate the old assumptions.

I expect the Amazon / Macmillan negotiation soap opera will resolve itself before long, but is in danger of drowning out discussion of the lasting shifts that the industry will inevitably go through. Here are a few of my predictions:

More competition with out-of-copyright works

In a physical book shop works that are out of copyright tend to sell for almost as much as recent releases. This is not unreasonable, since they still require editing, desigining and promoting, not to mention the cost of printing the physical book. Shelf space is a major cost in high street book shops, and costs no less when the work is out of copyright.

In a market for electronic books, some of these costs go away. The rest can be defrayed over a much longer period of time, since there is no opportunity cost to keeping books on the shelf. Competition with amateurish free copies will drive down the cost of a professionally-edited version to far below the cost of a new work, while still allowing publishers to turn a profit.

Most authors are not blessed with the wisdom of Marcus Aurelius, the wit of Jerome K. Jerome or the inventiveness of Arthur Conan Doyle (just three authors I’ve enjoyed for free on my Kindle). Of course the modern world has its own share of geniuses, and even those of middle rank can offer great value to readers by writing about the subjects that interest them, but it would be a sad world indeed if people didn’t capitalise on the opportunity to enjoy classic works for peanuts.

More flexibility in length of published works

It used to be that a book had to be of a certain length to be worth publishing. This is going to change, since the fixed costs of printing have gone away. The only question remaining is whether people will consent to pay small amounts for small amounts of content. There is no rational reason why not, but the field of paid content isn’t a shining example of rational consumer behaviour to date.

More competition from the back catalogue

The ‘long tail’ effect, where stock can diversify as shelving cost drops to zero, can flourish in the electronic market as the cost of warehousing drops out of the model. No book need ever again be out of print. This is going to help some authors and hurt others: books that have fallen from prominence aren’t necessarily bad, and if priced keenly they might prove a worthy substitute for recent publications at hardback prices.

More competition from the ‘gift economy’

Once you discard the requirement that books have to be printed on paper to be consumed, the barrier to entry to the market is much lower. Vast amounts of text in blogs, wiki articles and social network postings become readable in exactly the same circumstances as published books.

People who infer that this will drive the price of all textual content down to zero are unjustified in their conclusion and hopefully wrong. A well-researched, professionally-written and carefully-edited document is worth substantially more than an amateurish one, and a rational consumer will be prepared to pay more for it. However, for the portions of our reading time (hopefully not all) that are idle escapism, it may well be that the difference isn’t worth caring about.

Don’t shoot the messenger

I’m not saying the above out of ideology, but because I think the trends are inescapable, in direction if not in extent. More important is what won’t change: people will still exchange money for items that are of value to them. The lesson from the music industry’s experience is that the publishers that embrace the new rules and figure out how to turn them to their advantage will prosper.

Incorporating advertising into a web offering is often seen as a no-brainer: free money that you would be stupid to turn down, whatever your business model. However, I believe this assumption holds insidious consequences that can be bad for producer and consumer alike.

Advertising solves nothing

It always surprises me how blind people are to the cost of advertising. If I watch a movie on TV, I might sit through 15 or 20 minutes of advertising content that I wouldn’t say I wanted to watch. How much yould you pay to get 20 minutes of your life back to spend on something else?

Advertising has costs on the producer as well, in terms of trust. Seth Godin describes the problem eloquently:

The digital world, even the high end brands, has become a sleazy carnival, complete with hawkers, barkers and a bearded lady. By the time someone actually gets to your site, they’ve been conned, popped up, popped under and upsold so many times they really have no choice but to be skeptical.

Basically, it’s a race to the bottom, with so many people spamming trackbacks, planning popups and scheming to trick the surfer with this or that that we’ve bullied people into a corner of believing no one.

In addition to the hidden cost, advertising doesn’t bring more value into the economy, it merely reallocates it. Advertising-funded services aren’t really being provided for free, they are being paid for by the people who buy products that have been advertised (the sellers having passed along the cost of advertising to the consumer).

There are two possibilities as consumer of advertising-funded material: either you are one of the customers of the advertised products, or you are not. In the former case you are paying just as much for the content as you would if you paid directly. In the latter case, you appear to be getting a good deal by getting “free” content that is being subsidised by somebody else, but there’s a hidden problem.

The problem is that if you aren’t the one who pays, the content producers have no reason at all to cater to your tastes. If your favourite TV show is cancelled or your favourite blog stops being maintained, you don’t have any cause to complain: you were jus a freeloader, who was getting by on the fact that someone else was willing to pay for the content you enjoyed.

Missed opportunities from advertising

So advertising isn’t all that it’s cracked up to be, but it’s still the choice of individuals. If both the producer and consumer are happy with their ad-funded arrangement, even if it’s not 100% efficient, why should I care?

I care because I think using advertising has externalities. The most obvious is that making everything free at the point of use hides pricing signals. The amount of ad money made by reading a 3-page article is roughly the same as made from a worthless advert-packed spam page. In fact, you might be less likely to click on ads while reading a worthwhile article. Certainly there’s no way to show that you value one web page 3 times as much as you do another, when it’s all free at the point of use.

Right now I’m writing an article about internet business models, but it’s to my economic advantage to include terms like ‘mesothelioma’ or ‘acai berry’. In fact, I could probably do better for myself by dropping the whole attempt to create original content and writing keyword-heavy blogspam. Would this make the world a better place?

One of the most annoying things is that we’re missing the potential afforded by dynamic pricing. The real problem with charging for content isn’t that its value is zero, but that the price is rarely right. Paying newstand price for an online newspaper is a rip-off, getting it for free is a bargain. Somewhere between those two prices is the true value of the content, if only we could find what it was. Using advertising is tantamount to abandoning this search and letting the advertiser call the shots.

“Free” may not be as important as people think

On the topic of free news content, The Economist noted the following:

A poll by Harris Interactive for paidContent:UK, a website owned the Guardian, finds that three-quarters of Britons say they would switch to an alternative free news source if their favourite website began charging.

So what? Are you telling me that if Mars bars were free, their sales wouldn’t drop by 75% when they started charging? Nobody suggests that Mars have an unsustainable business model by trying to charge money for chocolate.

Strategically, a small number of customers who care about your product enough to pay is better than a large number of uninterested consumers, all other things being equal. With advertising in its current depressed state, it’s quite likely that payment from 25% of readers would amount to more money than they could make from advertising to 100%.

The social web has changed the action of “surfing”, in that it’s now as much about the social trends as the physical links. We still follow HTML links, but they are so numerous we must rely on the wisdom of crowds to navigate our path through them via Twitter, Digg, Reddit and so on. Even so, the activity of passively consuming content driven by immediate gratification rather than a predetermined goal is alive and well in Web 2.0.

Surfing seemed to work quite well on the old web, where there was relatively little content and only those who really had something to say bothered to speak up. It seems a whole lot less appropriate now that the web is a sea of too much opinion and too little fact.

I was away from my internet connection for most of the Christmas break, but had my new Kindle to entertain me. The combination meant that I spent most of my time reading published books, in a much more focused way than if I were reading articles on the web.

When I settled back down to my PC after the break and opened my web browser, my automatic thought was “I’ll just see what’s on the Reddit front page”. Somehow I found the prospect depressing in a way I’ve felt before but never so specifically. I came to realise that surfing the web was like living on a diet of Pringles: individually tasty, but ultimately unsatisfying.

The new year is as good a time as any to commit to changing behaviour, so here goes: I’m going to stop using the web except where I have a clear goal to achieve in doing so. I’ll continue to use it for researching specific topics, and I’ll continue to read RSS feeds of sites that have consistently good content. I’ll let you know how it works out for me.

I won’t say too much about the physical device itself, which functions very well. The e-paper screen is a pleasure to read from. The slightly reduced contrast means it’s tough in very dim light, but the trade-off of being able to read in bright sunlight is well worth it. The inclusion of a qwerty keyboard is a real marmite feature, but on balance I’m glad it’s there.

What really excites me about the devices is its potential to change the way I read. In the past I’ve left a trail of half-completed books in my wake, as I’ve never had the right one with me. Sooner or later I’d forget where I was in a book or even that I was reading it in the first place. If nothing else, the Kindle works well as a way of centralising my collection and tracking where I am with each one. Could I do the same with a few cheap bookmarks and a large sack to carry the books? Kind of, but the point is that I never did do that.

The built-in mobile phone circuitry that supports “whispernet” (global wireless delivery of books) looks on paper like massive over-engineering, since I purchase books far less frequently than I sit down at a computer. However, it changes the nature of the device from a portable book collection to a portable library, and that starts to feel like something out of a culture novel (the series of science fiction stories set in a high-technology utopia.)

Whispernet really comes into its own for newspapers and magazines, which benefit especially from arriving with you fast and effortlessly, and from old issues not taking up space. At the moment the catalogue of magazines is small and many suffer from the lack of graphics, but there are already some good titles that are very keenly priced. If I could transfer my subscription to The Economist to Kindle I might never read a paper magazine again.

One of the most overlooked things about newspaper and magazine delivery is its potential as a monetisation platform. Because it’s linked to your Amazon account, you can pay large or small amounts for content with ease. It may seem odd to say that I like having to pay for content, but he who pays the piper calls the tune. If I want something, I’d rather pay for it than hope that some advertiser values my eyeballs highly enough to pay for it on my behalf.

Using an electronic format essentially kills the second-hand market: since data doesn’t get worn out like a paper book does, resale of e-books by customers would compete unacceptably with new sales. Therefore Amazon has no choice but to prevent resale from happening. I don’t have a problem with this, provided that publishers recognise it and price their books accordingly. Second-hand sale of a book (whether for money or swapping in an informal economy) helps to defray the high cost of acquiring new books (at least part of which is in printing cost, and cross-subsidising printing of unsuccessful books).

In this sense, a Kindle book is less valuable than a brand-new hardback, since it has zero resale value. In the case of back-catalogue books it is also competing against virtually free copies of the dead-tree version from charity shops or Amazon Marketplace, which further lowers the price I’m prepared to pay. I suspect that the publishers who realise this and price accordingly will benefit.

One last thing about copyright: I don’t like the Digital Rights Management (DRM) of the Kindle. DRM is the technology that binds a copy of a book to a particular physical device, preventing you from lending the book to a friend or using it on an e-book reader not made by Amazon. It’s not unduly restrictive in everyday use, but the prospect of buying my entire collection again if Sony wins the format war is a bit of a downer.

However, I think the best way to ensure DRM goes away is to back up my voice with my wallet. DRM on music is on its way out, and this is not because of the people who forcibly broke DRM, but because the iTunes store proved that there was a market of people who were prepared to pay money and play by the rules. Adhering to the spirit of copyright when safeguards were in place gave the industry confidence in a way that arguing over the letter of the law never did.

“Streaming is the future for listening to music. No more time consuming downloads and using up space on your hard disk trying to store all your music,”

This comes from Jesper Theill Eriksen, senior EVP at Danish telco TDC, quoted in an article here. I don’t know his background, but I’m assuming from the context that this is someone who ought to know better.

Let’s take the misconceptions one at a time. Firstly, far from providing “no more … downloads”, streaming actually causes you more downloading by making you transfer the song every time you listen to it. When you “stream” music, all that’s happening is that the file is being downloaded to your computer in the background. The same file in the same format travels across the same internet connection as it would if you downloaded it. The only additional step a streaming player takes is to discard the data after playing it.

Which brings me to the disk space issue. The cost per gigabyte of modern hard disks is tiny, and few people have a collection that takes up more than a couple of percent of the space on their disk. If you need more disk space, it can be purchased for far less than the cost of most streaming services.

As I’ve said in more depth here, streaming is nothing to do with convenience to the user and everything to do with trying to defend a business model: if the songs are on their server, they control them rather than you, and have more leverage to demand money from you on an ongoing basis. Myself, I’m happy to let the market decide, but let’s see streaming for what it is: a business model that operates by taking away your ability to control the music on your devices.

1. Every single non-alphanumeric character in the ASCII set is put into a hat. Players draw from the hat without seeing what they are drawing.
2. At the beginning of the game, LISP draws two matched parentheses from the 2-card “LISP deck”, which is then exhausted. Both LISP and the LISP deck play no further part in the game.
3. The players sit in a circle, and play proceeds clockwise. Programming languages choose in a random order, apart from ALGOL, who chooses first.
4. Each player borrows 3 characters from the player to their right, and draws from the hat as many additional random punctuation characters as they feel they need to create an expressive language and a related documentation format. They must then choose an escape character, which must not be the same as any of their characters, but must duplicate a non-escape character picked by a previous language.
5. At any point during the 1990s, a player may play the Unicode card. From then on, the original ASCII hat is supplemented by a Unicode bin containing every single character used by a living language, and most dead ones.
   • NB: It is a common house rule that all players ignore the unicode bin with the exception of F#, who chooses last and often has to root around in it for some unused characters
6. Any player who feels that the game is proceeding too slowly may play the SGML card, and may pick any matched pair of punctuation used by a previous player and redefine it to be the foundation upon which further markup is based. Subsequent players no longer need to define a documentation format, though they must now pick two escape characters.
7. Once all the players have chosen their punctuation, each must write an operating system. Points will be deducted for any code that is valid in more than one language.

The recent fuss about micro-lending site Kiva not being all it appears to be reminds me of Paddington. A certain segment of the internet community is shocked, shocked, to find out that the loans they’ve been deciding whether to make or not have already been signed, sealed and delivered in the field. For people who were assuming that they had the power to decide who did and didn’t get a loan, this is undoubtedly a surprise.

The situation isn’t as bad as it has been made out to be: I’m assuming there is some sort of feedback effect between loans made and future loans funded. If Kiva lenders decide en masse to support only women, or goat-farmers, or people in Cambodia, then Kiva will request more such loans from their field partners and eventually more such loans will be made. Nor is it the case as far as I can see that Kiva is double-counting loans to a recipient. If you want to believe that the dollars received by a particular Cambodian goat-farmer are in some sense “your” dollars, then you can continue to do so.

Like Paddington, some charity donors seem to be confused about the fungibility of their donation. Charities may promise that your donation goes to the project of your choice, and maybe even that none of it is spent on administrative costs, but this is of course a half-truth. If your donation is added to the total on Project Y, this may mean that somebody elses (unrestricted) donation isn’t needed for Project Y, and that that donation can be used for administrative expenses. The net effect is no more benefit to Project Y, but a decrease in the deficit for administrative expenses.

This is as it should be. Administration is a vital link in the chain that drives the quite remarkable (though often unremarked-upon) process of turning figures in my bank account and clicks of my mouse into clean water and education and food for some of the poorest people in the world.

Furthermore, SEO isn’t hard because it’s a conceptually hard task (which would have intellectual respectability), it’s hard because Google and others have deliberately made it hard. The search engines publish far less information about their algorithms than they could do, because security by obscurity is much easier than creating an algorithm that can’t be subverted.

A further thought occurred to me more recently, based on the idea of “rent exhaustion”. This is a term that has been thrown around on a few blogs (notably Tim Harford’s and Tyler Cowen’s) and is allegedly a mainstream term in economics, though I haven’t found any authoritative definitions of it. In any case, the example usually given is this: handing out money to beggars doesn’t increase the overall wealth of the community of beggars, it just causes them to waste time and effort crowding round for handouts. Assuming a beggar’s time isn’t worthless, any time they spend queuing for a donation represents a net loss to society.

I wonder whether SEO doesn’t represent just such a waste of resources. The “handout” in search engine terms is the large slice of customers’ cash that goes fairly indiscriminately to whoever is at the top of the Google rankings. The more customers there are, the more effort is expended on scrabbling for their attention. Customers aren’t being served by better products at keener prices, and they’re not even benefitting from attractive and exciting marketing (which we may enjoy more than we care to admit).

Is there a way to resolve this? A perfect search engine could distinguish good content from spam, but doing this is probably a strong AI problem. Perhaps search consumers can make the world a better place by venturing beyond the first result page more often, and critically evaluating the sites they are offered, but it’s not clear that this highly-duplicated effort doesn’t represent a waste of resources in itself. Ultimately a solution might be to pool the opinions of savvy consumers to contribute to the ranking, which was of course the original goal of using link data as a proxy for recommendation. Until we solve the problem of identity on the internet, any such system will be exposed to manipulation, leaving the opinions of real consumers a vastly under-utilised resource.