Business Models

Interesting story today on RADAR about Gawker Media cutting revenue per impression for its hacks even while revenue and traffic increases:

....the per-employee traffic isn't that much higher than it was a year ago. And yet the site traffic is up more—meaning the site is receiving more income that the company doesn't have to share with a writer. The site received 16.7 million pageviews in June. Only about 6 million pageviews of that traffic is attributable to writers currently being paid. So why is the company cutting the costs of staff pay, when it isn't forced to cut in writers for 10 million pageviews in the month?

One of my frustrations with many of the journalists covering the new media / tech space is their very cursory understanding of economics, which means that all sort of hype and cr*p gets picked up and reported on uncritically, so it is with a certain wry amusement I read this

The issue is that unlike print where yellowing copies of last weeks' newspapers are good only for fish and chip wrappers, Digital Media is persistent, ie that stuff written 10 years ago by people long gone is still garnering traffic, and as time goes by this rises - as shown in the diagram below:



So, in a CPM based Ad serving model, as time goes by the proportion of money a site earns from current output reduces, while the money earned from existing back-catalogue output increases - so that after a while by far the majority of its income is potentially coming from media already written. (There is a "half life" to old material, but if you model it you find say 5 years of Old stuff has a large presence at fairly low visit levels - you are essentially building a larger and larger "long tail" )

At this point a bit of game theory analysis does not go amiss - given that the historical content is lower cost to the owners, as they don't have to pay out the original writers (assuming all content rights belong to owners and there is staff turnover over time ) then the reliance on the existing writers diminishes. Those who pay the piper call the tune, but if you don't even have to pay the piper 'cos you've kept the tune.....

For writers to reduce this tyranny of persistence they either (i) need to keep the right to their own historical material (unlikely, as per page payments would reduce) or (ii) current writers must make contracts that take a share of historical traffic (harder and harder to do as reliance on them reduces). Otherwise its this way to Digital Sharecropping.....

There is a second impact of Ad based media paid by the page, which is that by definition "populist" stuff is better rewarded - so expect more sleb slavering and political polemics - and niche stuff is not, so for example balanced reporting on complex issues is a route to starving in the garret.

So anyway, next time all you online hacks wax lyrical about FreeConomics or that the New Media defines a New Economic Paradigm, take a look at your own pockets and note that same New Economics is thinning them out via some very Old Rules, that was well understood in that most deeply unfashionable of places, the Main Stream Media.

Oh, and you may want to look up a guy who sussed a lot of this about 150 years ago. Name of Marx......

(Update - another take on it here by Jason Calacanis - expands on the above thoughts quite nicely....)

The perennial “Spot the Twitter Business Plan” game took an interesting turn today. There is an argument on Silicon Alley Insider, the thinking being that its Universal Comms nature and simple transaction messaging structure makes it ideal for a multi-media small value payment service.

If Twitter had a P2P payments system in place today, it would become the most used mobile payments system overnight. Having the ability to send a message like “p(ay) @broadstuff $5″ for that beer I just bought you would integrate seamlessly with the way Twitter’s users already interact with their system.

Layering on a payments system would not only make the feature instantly used, it would position Twitter to revolutionize how money is collected and exchanged on the Internet. (Think of what Twitter’s done for flashmobs and how it could effect fundraising.)

The argument is also the Twitterers are well used to using coding to enact various transaction messages, so why not transactions too:

One of the most missed facts in the mobile payments space is that users of a system have to be comfortable communicating using machine language. This is to say, one must remember and follow certain semantics so the system knows how much you’re paying and to whom.

Twitter users are already trained in this important action. Every time a Twitterer uses the “@,” “d” or even “#” to direct Twitter or annotate the messages it sends through the system, people are using the exact sort of machine language they’d need to use for mobile payments to work.

Interesting idea – the obvious thought is to link it to Paypal. That way eBay could integrate it with Skype as well and maybe finally create some value there.

Even more interesting that the concept of actually paying for extra services is still unmooted - judging by the wails when it goes down, you would have thought a small fee for say 100+ followeds would not be out of the question?

Update - comment by Mike Dopp via Twitter:

Wouldn't say I would pay for it. However I would not mind google ads or others to pay for hosting and bandwidth costs

In hindsight I think this is on the right track.....

There is an interesting discussion going on on Slashdot - Chris Anderson's long tail being pulled by an HBR article potentially debunking his Long Tail thesis, (update - its made the WSJ now - this one hurts) though Chris's view is that one person's head is another's tail:

The best example of this is in what she describes as a growing "concentration" of sales around a relatively small number of blockbuster titles. In the Rhapsody data, she finds, the top 10% of titles (out of more than a million in that data sample) accounted for 78% of all plays, and the top 1% account for 32% of all plays. That sounds pretty concentrated around the head, until you reflect, as she notes, that "one percent of a million is still 10,000--[...]equal to the entire music inventory of a typical Wal-Mart store."

Actually, the Long Tail has been around for as long as inventory has been stored, the basic maths was done in the last century in the early days of statistics (Pareto ring a bell?). The obvious points made are that:

The whole premise of the Long Tail idea is that it's NOT a hassle for internet companies like Amazon and Apple to keep the low-volume stuff in inventory. That's what gives them an advantage over brick and mortar. If Amazon only sells 5 copies of the The "Long Tail: Why the Future of Business is Selling Less of More" a year, it only needs to keep 5 in inventory. If a chain store wants to sell the book it has to keep hundreds in inventory to ensure they have at least one in each store.

A lot of people mis-characterize the Long Tail as "making money by selling obscure stuff." That's only half of the definition. The other half is building a business where it does not cost you anything to keep lots and lots of obscure things in inventory, or alternatively, having access to millions and millions of customers, so if you sell to 1% of your market, it still adds up to something. Long tail does not work unless all of those 1% niches add up to 80% of the total market or unless you sell to 1% of a billion people. The internet helps businesses do both of these things.

There is one real concern I have of Long Tail theory, which is that another impact of low transaction cost networks is you get positive return dynamics, ie the rich get richer - faster - and I've not seen a convincing argument as to why this dynamic doesn't occur as the endgame in any Long Tail initial condition system.

This discussion however was my favourite:

(Person One) But then try to explain porn websites. There is a lot of tail to be hit there.

(Person Two) And lots of head too...

I love Slashdot discussions, there is a rigour and sharpness, as well as irreverence, that I think is missing in social aggregators such as Friendfeed. I'm not sure why......

Interesting piece in TechCrunch re valuation of Social Networks, the really interesting bit is this table:



Now we have had a look at these valuations before, (see here and here for example) and our view is that the tiny % investments in Facebook and Linked In don't really show value, whereas outright takeouts like Last.fm and Bebo do. Thus the column from Bebo's valuation are in our view more indicative of the real values - especially given that there is increasingly a view that the early assumption of Ad revenues are overblown

There are clearly conditions to the valuation, and this was a good comment from Rich Swier:

....but from a pure valuation perspective, you are missing some key elements like

1. Dead Profiles - The biggest “dirty little secret” noone talks about is how many of these sites have dead profiles, where people never login again. If you look at Alexa (etc) you can predict how active the site really is. I think most of these sites have a serious retention problem.. especially Facebook (once kids graduate and become adults)

2. User “willingness” to Pay - this to me is the biggest factor you overlook to some degree. Its great to have millions of users, but if they are not willing to pay for premium features, then it will be hard to monetize. Even the biggest networks make very little money on ads… (because users are there to socialize not click on ads (e.g. shop). To me this is the key to a valuation. If you can’t make money? What’s the point?

3. And finally, the other factor to Valuation is “who will buy it?”… I don’t think any of these companies can or want to go public.. so who is the buyer? if there is no buyer, then the valuation is meaningless…. I think sites with critical brand recognition like Linked In in the pro-market, will have a longer list of buyers…

All very true, but given that most of these networks are chasing similar (yoof) demographics we can probably assume 1 and 2 are more similar than different (Linked in being an exception) Point 3 is always key - there is a limited set of buyers for these services.

I'll be giving a short talk tomorrow at a panel on the BBC Media Futures event on the Myths of FreeConomics. That'll be the 5 minute version, for the full half hour start on the blog over here. If you do read this blog, come over and say hello.

Its in the "Provocations" stream at 3pm

So, the AP has now riposted to The Brouhaha with a re-post of its pricing for quoting its words (see here for the price tabs):

There are 3 rates - For Profits, Educationals, and Registered not for profits. The Educationals and Not for Profits have the same pricing. I will now perform my own value-added analysis, which last time I looked qualifies as Fair Usage. If you graph the rates, it looks like this:



As you can see, there are a few interesting points of inflection:

- The pricings are the same, no matter what the profits, for 51 to 250 words - clearly this is the desired sweet spot
- There is an arbitrage at each of the breakpoints, typically of c 2x the following word valuation

Reflecting on this inflection, we note that the end-point converged value per word is c $0.20 (the graph tends to this over and over), and it takes c 400 not-for-profit or 500 profit words to get there as a larger than 250 word buy, which we submit is the lower bound of the assumed "Fair Use" envelope

This leads to some interesting conclusions:

- the average page in a book is c 300 words, or $60 at $0.2 per word. Assuming the average book is 300 pages, that values the latest schlockbuster novel as c $18,000. Now, given that we want to pay about $9 or so, that implies a c 50,000% discount for a bulk buy of these novel words.

- assuming the average Sunday newspaper is (say) 500,000 words, and sells for £2.50 (say $5.00) that implies a price of $100,000, or a discount of 0.5 million % to reach the price of the papers in the shops. The photography of course is free in this calculation, but assume thats half the cost and its a mere 0.25 million % discount.

- assuming the average newspaper article is c 500 words, thats' $100 an article. Assuming the average Journo makes $60,000 per annum, thats c 600 articles a year, or c 2.5 each working day before you are cheaper than buying straight off AP. And thats c 5 a day if they get a 50% discount at your paper, bub, so stop reading this blog and get writing .

Anyway, as we can see, AP is clearly selling its words at substantial discount in bulk within the trade compared to the Fair Usage options.

(Afterhought) I don't think I know of another market where these sorts of pricings apply, especially one for digital goods which have low intrinsic distribution costs.

This post comes from something thats been rattling around in my head awhile, which is this - if the wisdom of crowds is so great, how come Pandora, with its analysis based "music genome" algorithms, gave me a more consistent selection of music I like compared to the crowd-sourced Last.fm?

(By the way - huge kudos to Improbulus for showing how us UK based musophiles can get Pandora again via her article on Geospoofing - we luvs the Internetz )

If I look at my experience of the key differences between the two, it is this:

(i) Output quality - LastFm, like Amazon, selects based on what others who like X also like. So for example, if I like the Sex Pistols and Abba, and you love Abba - guess what - at some point a crowdsourcing system is going to serve you Anarchy in the UK. Last.fm now does have a "music like X..." option, and I've been playing with it over the last 3 months, but it just does not optimise like Pandora can. Both systems allow you to like / not like, but I suspect it is just harder for a system where the base data is more flawed to optimise as well.

(ii) Input Quality - crowdsourcing systems depend on people putting in what they like - which, in my experience, means a shorter tail, its not just me who has noticed that Pandora serves a greater variety of acts that accord with your taste as its not as bothered with popularity of any kind

Now I did notice when I started last year that Last.fm got up and running to converge around an optimum faster, but once Pandora was on the hunt it quickly overtook Last.fm in quality of the experience. Also, I suspect that its harder to "game" Pandora wth SEO link love etc than any crowdsource based system, which is prone to various schemes to artifically boost perceived popularity.

The impetus for blogging all this now is that I have recently finished a rather interesting book, Super Crunchers, which is in essence the emerging story of data crunching in a high data web 2.0 world with very cheap megaflops. The implication of this book is that its is now both operationally and economically possible to crunch extremely large datasets and derive subtle links that no "wisdom of crowds" system could achieve (as I'm sure Google et al know only too well).

Also, as anyone who has studied social networks knows, there are some quite difficult precondions that have to be in place to ensure you get a "wisdom of crowds" rather than a madness or groupthink of crowds going on. The minute we see feedback, we stop thinking for ourselves and drop into socialthink mode. It is extremely doubtful if these preconditions exist in any current social nets, or even if they are possible - and if they are, whether they are stable as they are prone to gaming.

Thus, if there is a claim for a "Web 3.0", where "Web 2.0" is the interactive web, its the emergence of the "supercrunched" Web (which also plays in to another trend, the rise of the m2m web I think). And there is a lot of storage and crunching capacity being built ( And just who is building many of those datacentres - ah - Google. Funny that )

There are going to be downsides - privacy for example. I don't know what Pandora could deduce from me by knowing my taste in music, but I suspect a psychologist could have a field day - and if it were linked to my activity elsewhere I suspect - as behavioural psychology and economics is showing - that it may know more about me than I do. Thus there will be a very strong argument to keep various niches of our data separate, for us to optimise our experience in vertical.

By the way, I suspect that the endgame will be hybrid systems, and that a Social Media system may (may!) give a higher level of serendipity simply by the nature of how its architected

But this Pandora's box, just like the first one, once opened, can't be shut again.

Now, I'm sure the criticism will be levelled that I have just succumbed to the "groupthink" mode by allowing myself to be seduced by this book, and that in fact no algorithms can reproduce the subtlety of a million, a billion human minds. My counter - that's possible, but the evidence, in my view, sings for itself

El Reg notes that Google has now said it will give users the tools to see if bandwidth is being throttled:

We're trying to develop tools, software tools...that allow people to detect what's happening with their broadband connections, so they can let [ISPs] know that they're not happy with what they're getting - that they think certain services are being tampered with," Google senior policy director Richard Whitt said this morning during a panel discussion at Santa Clara University, an hour south of San Francisco.

If the country doesn't have neutral networks, Whitt contends, innovation stagnates among application developers. And he believes that individual consumers - as well as Washington policy makers - should join the fight for such neutrality.

Which always brings me to the bigger question of "what do we mean by Net Neutrality"?

By this, if we mean that we should not constrain one particular bit over another because we don't like what it does, I'm all for it - that bit of "all bits are Free" I get.

If we mean that all services should have equal opportunity to access the 'Net, and ditto all users, then I'm your man for that bit of the Free Net.

If it means that all bits are free, as in not priced, then the Tragedy of the Commons leaps to mind - for those not familiar with this effect, it is this - imagine a patch of common land, where all herders have rights to graze their animals:

The herders are assumed to wish to maximize their yield, and so will increase their herd size whenever possible. The utility of each additional animal has both a positive and negative component:

Positive: the herder receives all of the proceeds from each additional animal.
Negative: the pasture is slightly degraded by each additional animal.

Crucially, the division of these costs and benefits is unequal: the individual herder gains all of the advantage, but the disadvantage is shared among all herders using the pasture. Consequently, for an individual herder weighing these, the rational course of action is to add an extra animal. And another, and another. However, since all herders reach the same rational conclusion, overgrazing and degradation of the pasture is its long-term fate. Nonetheless, the rational response for an individual remains the same at every stage, since the gain is always greater to each herder than the individual share of the distributed cost. The overgrazing cost here is an example of an externality.

In 'Netland this means video (and usually P2P video) - video users consume a LOT more bandwidth than non video users. In essence the majority of 'Net users today do not use video heavily, so are subsidising those few who do. But where the rubber hits the road is when the total demand for the bandwidth starts to come to the point where the pipes are nearly full, and someone needs to pay for the upgrade.

Who pays - why, the ISPs and Telcos of course, say the Net Neutrality crowd.

Except (i)...imagine you are running an ISP and find you have to spend billions to upgrade your pipe just so a fairly small number of superusers can carry on pumping moving pictures around at no extra costs to themselves..... and that that small number will continue to grow as there is no penalty to them doing so. Infinite demand, infinite cost, very finite revenues.

Except (ii) Those pipes aren't commons - they are owned by private operators, and those operators have 2 choices - (i) make a profit or (ii) go out of business. And their obligations to shareholders says the latter is to be prevented.

Ergo, there are only two realistic options - prices must go up, or over-users must be throttled.

Now its easy to see why Google doesn't want this to happen - the economics of YouTube and other new high volume services are based on this subsidy of the bandwidth rich by the bandwidth poor. To be fair, till now ISPs haven't helped themselves as they have flat priced bits while the pipes were empty, thus setting expectations. But is this sustainable - can you go on adding traffic to increasingly constrained pipes and expect the owners to upgrade at no benefit to themselves? If they remain private companies, the answer is no. And one can look hard at the Net Neutrality argument and say that one "gets" the argument for freedom of all bits, but not for free as in zero cost bits. Different issues.

It is for this reason we suspect / believe the Net Neutrality argument, to be credible going forward, needs to purge itself of the "Freeconomic" element of its thinking, as that is not sustainable as demand rises. In nearly every constrained resource in history, typically at some point there is a payment system implemented for use made - or else it tends, as Jared Diamond notes, to collapse for all:

He cites five factors that often contribute to a collapse, but shows how the one factor that all had in common was mismanagement of natural resources.

Natural being common, mismanagement being unsustainable use...which, in our case, is bandwidth.

This article in the WSJ crystallised some thoughts I've had recently - it was about male vs female styles in code writing, and the paragraph below struck a chord:

There’s a big need to fix testosterone-fueled code at Ingres because only about 20% of the engineers are women, McGrattan says. (Most of them are in jobs involving quality assurance or adapting the product to a new locale, she says, and not the “heavy lifting” of writing code.) She’s on a mission to get more women interested in computer-programming careers. But “it’s proving very challenging,” she says.

This is interesting because I recognise the issues, knowing many women in IT and having been at and run software companies, and I take my hat off to people like Sarah Blows and Co who are driving the issue with Girl Geek Dinners in the UK, and others doing it elsewhere. Also the number of good girl geek bloggers must help.

However, this is also the direct antithesis to what's happening on the user/customer side in social media. Women are now the majority of people on Facebook, and have high representations on such services as Second Life, Twitter, Bebo, MySpace etc. I have also noticed a lot of women are involved in the various "Social Media" style "geek" events such as Social Media Club. Web 2.0 has a much higher female composition at the user end, anyway. Women are also driving online retailing and the evolution of a number of online services, such as gaming.

Hence my thought that social media services have feminised the Web to a level not seen before.

Talking to (female) friends of mine about why this may be, I've had some interesting insights - among them:

- On Twitter, having pictures of people makes searching very easy compared to text based systems such as email
- What pictures people choose to put on their icons and profiles is revealing
- Ditto the data they put up on profiles
- On Plurk, the timeline is more intuitive than previous linear systems
- Its very useful to see who someone is friends with that you also know - allows you to "place" them better
- Video allows a much higher bandwidth conversation (conversely, more worried about appearance when seen)

The impression I get is that Social Media allows far more context about a person to be seen very rapidly, which is clearly very attractive to women. (By and Large) women are not turned on so much by the technology per se, but how you use it. Most social media is used for communicating (the killer app for women and men as well) and is easier for womn to use than previous stuff. One can theorise why this all is, but regardless of the theory, its a fact.

Also, I've been chatting to a number of smart women about how the web may evolve to provide women with services tailored more to them (given that they control more household spend than men), and the differences in insight and nuance are quite striking, and when taken together signal big changes in how you may define what the service does, its functionality, user interfaces, service design etc.

And thus, one would think, that successful companies in this new space may find it very useful to have women embedded at all level in the process of conceptualisation, design, development etc - it must save money and increase the chance of the service succeeding.

So getting that number up from 20% would be a very good idea, shirley?

(Update - this article on CNet puts a window on another issue - women in geekwold don't always help themselves here.....)

Article today on Business Week (saw it on Nick Carr's blog) re eBay auctions falling off:

Bruce Hershenson, who auctions vintage posters online, is hanging up his eBay gavel. For almost a decade, Hershenson's business epitomized the e-commerce that made eBay (EBAY) famous. He sold rare, collectible, sometimes kitschy memorabilia in online auctions that had a starting bid of 99¢. But as the business of buying and selling over the Internet has matured, the thrill and novelty of auctions have given way to the convenience of one-click purchases. Hershenson will hold his last eBay auction June 3. "The auctions are nothing like what they once were," he says. "They won't ever come back."

Auctions were once a pillar of e-commerce. People didn't simply shop on eBay. They hunted, they fought, they sweated, they won. These days, consumers are less enamored of the hassle of auctions, preferring to buy stuff quickly at a fixed price. Hershenson is emblematic of the legions of small business people who built their livelihoods on eBay but—like eBay itself—are having to rethink their whole approach to online sales.

Sad perhaps for eBay, but fascinating for armchair digital economists

I think there are 2 opposing forces going on here - the (higher) transaction cost of an auction vs the probability of achieving the (lower) cost one may get the goods for.

Firstly, I think we are getting much more experienced at judging when to use the auction and when to use the "buy now" facility. Set of drums - big saving, worth it - drumsticks or a cymbal - too much hassle.

Secondly, I think that eBay has not been managing the creeping degradation of the auction service experience for the average user - if you ever take part in an occasional auction now, you know how much cuter the others are than years ago - some are clearly using gaming software, or are just very experienced - which just makes you go away in frustration.

However, I also think this is partly a result of the auction's success - the fixed price retailer can't put too high a markup on the goods, or else people will opt for the auction. If auctioneering dies away, it would be fascinating to see if (i) the price spreads increase and then (ii) Auction traffic comes back.

And Nicks view of eBay's future - he notes:

EBay made a ton of money running auctions over the past ten years, and it may continue to be successful as the operator of an online mall. But it is not the company we imagined it to be. Its story has become a cautionary tale about the dangers of wishful thinking and fanciful extrapolation.

I think having a market the scale of eBay's was - and is - a huge source of competitive advantage. Getting the transaction economics and experience right will be critical as the mass market comes on stream.

Oh, and spending money on things like Skype - maybe they should set up a predictive market when they want to do stuff like that