broadstuff

Proof that hope springs eternal in the human breast, this article on Twitter being a spam free zone is yet another example of touchingly naive optimism (thanks for link via Howard Lindzon - via Twitter of course).

Sez Russel Beattie:

Think about any other online community system ever created - from Usenet to The WELL, IRC to Slashdot to Digg. All of them have had to deal with the core problem of idiots on the Internet. Slashdot's extended karma system, or IRCs multiple moderation commands, Digg's diggs, etc. are all about filtering out the assholes. It's a fact - anytime a virtual group gets to a certain size, the morons come and start doing their best to disrupt, defraud or degrade the audience.

Twitter, however, has almost a million members, a thriving community, lots of discussions and yet doesn't have spam or troll issues. If you step back for a second and think about it, that's pretty amazing.

The fascinating thing is timing - last night a few people were starting to talk about the rise of spam on Twitter. Right now we can see 3 types:

(i) "A List" Spam - some influential bloggers use Twitter as a broadcast system to blog (blag?) about every post, activity etc they are involved in. Much of this is purely an attempt to sell (flog) their wares.

(ii) "Z List" Spam - people who have such a high post rate that you just have to turn them off

(iii) And now, the rise of Twitter Spambots - this was what we were discussing last night, ie a number of people are seeing spambots (or spampeople - on Twitter its hard to tell) who tend to follow huge numbers of people but have no following themselves, and are probably scraping data all the while, and (the topic of conversation last night) some people are starting to see the first Spam swallows of a Spam summer.

Sadly, PR/marketing spam kills most of the things we love once it they get big enough to be worth going after, and Twitter's privacy controls etc are still so rudimentary (You have to turn off all @ comms for example), that it is fairly easily spammable - it just hasn't been worth doing it until recently, but now its getting big, spam will most surely come.

...was when I saw this article on plants using it to call for water. Why? Simply because our research shows that the messaging from m2m (machine to machine) applications will dwarf human messaging. The sheer number of devices that are going to come online in the next few year in digital homes, businesses, vehicles etc will drive a step change in internet (and mobile) usage. We also believe the main early instances of the semantic web will primarily be to facilitate m2m comms.



Besides, its probably far more interesting content than many people put on Twitter anyway - and plants don't spam you.

In rapid succession, Fred Wilson and Howard Rheingold used Twitter to communicate articles they had just written on their respective lurve for Techmeme and Twitter.

First Fred, on A VC blog:

Seeing the big news headlines of the day is important and I love watching them come through in my twitter feed. But seeing the related links and knowing when someone you trust is talking about an issue is the big deal for me in Techmeme. Others do this as well, but to my mind Gabe does it best and that's why Techmeme is a daily read for me.

My view exactly - the killer app for Techmeme is catching all the commentators as well as the A Listers, adds depth and width. The risk Techmeme et al run is that that "Big Story" board becomes too self referential (ie only takes stories from newspapers and A - Listers - see our earlier post on Digg in this respect - and its interesting to see how the Techmeme input streams have concentrated over the last 4 years), but that is mitigated by the commentor model.

Howard waxes eloquent on Twitter in Smart Mobs, on the benefits of Twitter:

Openness — anyone can join, and anyone can follow anyone else (unless they restrict access to friends who request access)

Immediacy — it’s a rolling present. You won’t get the sense of Twitter if you just check in once a week. You need to hang out for minutes and hours, every day, to get in the groove.

Variety — political or technical argument, gossip, technical info, news flashes, poetry, social arrangements, classrooms, repartee, scholarly references

Reciprocity — people give and ask freely for information they need (this doesn’t necessarily scale or last forever, but right now it’s possible to tune your list — and to contribute to it — to include a high degree of reciprocation)

A channel to multiple publics — I’m a communicator and have a following that I want to grow and feed. I can get the word out about a new book or vlog post in seconds — and each of the 1300 people who follow me might also feed my memes to their own networks. I used to just paint. Now I document my painting at each stage of the process, upload pix to flickr or flicks to blip.tv, then drop a tinyurl into Twitter. Who needs a gallery or a distributor?

Asymmetry — very interesting. Very few people follow exactly the same people who follow them.

A way to meet new people — it happens every day

A way to find people who share interests — I follow people I don’t know otherwise but who share an interest in educational technology, video, online activism.

A window on what is happening in multiple worlds, some of which I am familiar with, and others that are new to me

Twitter is something I am less certain of - it certainly is a good broadcasting channel for people with big followings, but I would treat with a pinch of salt any of these people's claims on reciprocation. The experience of being on the "receiving end" is all too frequently one of being a consumer - in fact I have turned off quite a few "A Listers" as their post frequency and unwillingness to converse makes their Twitter contributions akin to spam sometimes.

As with any conversation system, it is easier to meet new people than via totally asynch systems, and as with any group network based system it get a wider variety of inputs - so long as you sign up to a wide variety of different people.

One emerging good point about Twitter is that the "River of Drivel" is declining, as people now spend less time letting everyone else know about the banalities of their lives, and start to self-edit for decent content.

The other point that does interest me - and goes back I think to Fred's point on Techmeme - is that there is asymmetry in the following / follower area. I think there are 2 reasons for this:

(i) You follow some people who are interested in the same stuff you are, but the vice is not always versa.

(ii) At present, we all have different rules for following people who follow us, I suspect this drives a lot of the asymmetry.

However, its not as efficient as the older email/groups systems for message persistence (ie if something was said yesterday you're unlikely to go back and read it today) or message management (eg sorting stuff, filtering irrelevant posts) nor does it have the immediacy of real time chat.

What it does have over the Web 1.0 chat systems is the real time view of "friends of friends", and as we've noted before this is the slightly voyeuristic (and thus often unstated) attraction of today's Social Nets. In addition it is a good UM system as it can be sent and received across so many devices.

But I do feel that the endgame for Twitter et al is the blending of the functionality of Groups and email with the voyeurism of FOAF, and thus I am less clear on its future - that and the fact I haven't got a clue how it will make money apart from banner ads on the website.

Very interesting article in Slate on refining the 90:9:1 rule of Social Media. There is a fascinating paper on the development of the Wikipedia contribution ecosystem referred to there.

Slate notes that, re digg and Wikipedia:

While both sites effectively function as oligarchies, they are still democratic in one important sense. Digg and Wikipedia's elite users aren't chosen by a corporate board of directors or by divine right. They're the people who participate the most. Despite the fairy tales about the participatory culture of Web 2.0, direct democracy isn't feasible at the scale on which these sites operate. Still, it's curious to note that these sites seem to have the hierarchical structure of the old-guard institutions they've sought to supplant.

This top-heavy structure of social-media sites isn't news to researchers and technophiles. Wikipedia co-founder Jimmy Wales has acknowledged that what he expected to be an "80-20" rule—a system where 20 percent of people control 80 percent of the resources—in fact understates the site's top-heaviness.

Whereas for Slashdot..:

The moderation system at the tech blog Slashdot is perhaps the best example on the Web of a middle way. Slashdot, which draws on links submitted by readers, ordains active contributors with limited power to regulate comments and contributions from other users. Compared with Wikipedia, which requires supreme devotion from its smaller core of administrators, Slashdot makes it easy to become a moderator. Giving large numbers of people small chunks of responsibility has proven effective in eliminating trolls and flame wars in the comment section. Still, the authority any one moderator commands is small, and the site's official poobahs maintain control over which stories are featured at the top of the site. "These things are far from utopian," says founder Rob Malda, aka CmdrTaco. "Slashdot tends to have a lot of 'Microsoft does something bad' stories. If I let the community run the whole thing, we'd have a lot more.

What is interesting though is that the Oligarchies are slowly becoming more democratic ( a trend we also suspect is happening in blogging ) - here for example is the chart from the paper for del.icio.us



The paper notes that:

A key difference between del.icio.us and Wikipedia is that the former does not promote direct interaction between users; instead, its power derives from the aggregation of many users’ individual data. As such it is an interesting contrast case to the high degree of interaction found in Wikipedia. We examined the distribution of work over time in del.icio.us as measured by the number of bookmarks added per user. As in the earlier analysis, users were split into classes based on their total number of bookmarks added. (The Figure above) shows the percentage of bookmarks made by different user classes. As in Wikipedia, we see a marked decline in the percentage of edits made by the highest-edit class from a high of 78% to a low of 27% in the latest data (June 2006). There is a corresponding rise in the lowest-edit class, from a low of 3% to the current high of 31%. Note that del.icio.us shows only a steady decline in the influence of elite users, with no initial rise as seen in Wikipedia. This is an intriguing difference that merits further study.

Rory Cellan-Jones writes today about the decling Facebook population on the BBC Technology Blog

Facebook - it's so over. That's been the tenor of most of the commentary since Thursday's figures showing a slight dip in Facebook's UK users. The general feeling is that the kids, with their minute attention spans, have already tired of the social networking site and moved on to something more hip and happening. I think the opposite is true - that Facebook's new wave of older users have decided it is just not worth the bother and are now leaving it to the kids.

Thats possible, we've noted before that the mavens are leaving Facebook (and appear to be going onto Twitter) - but it could be something much more mundane, ie just a slowdown in growth, which is never smooth.

But I wonder if most people understand what this slowdown means for something like a Social Network, where usage is heavily front-loaded.

Consider the chart below - of the 100% of time most people spend on any one social network, most is in the early days - setting up the friends, playing with the features etc - and then erodes over time. Below is a theoretical graph, showing an average halving of activity every 3 months over the 2 years average time that a person exists on a social network in any meaningful way. (Note - I don't mean people necessarily leave after 2 years, just that activity - on average - is fairly low. You can make less extreme power laws if you like, but the result described below is much the same until a fairly low tailoff rate is assumed)



Now imagine how that curve works in a Social Network that grows from next to nothing to say 50m users over 4 years until growth tails off. As you can see, the traffic initially grows far faster than the new user growth, so its boomtime in impressionsville....but as growth starts to slow down those users' reduced activity starts to kick in, and the great crash in usage traffic then kicks in (see below).



"Usage" above is shown as an imaginary unit so the 2 curves are easy to juxtapose.....Users are cumulative users in millions.

Now clearly this is a theoretical model, and any real-life Social Network will be adding all sorts of features to keep the traffic going, but as this quick analysis shows, when growth slows down life starts to get very........ interesting.

(UPDATE - looks like a similar pattern is observable in the US as well.)

UPDATE 2 - now this is interesting, Facebook insists that unique visitors are still rising - which of course, is highly consistent with usage falling, following the discussion above - its just that they are not rising now as quickly as existing users' usage is decaying.

UPDATE 3 - its now Wed 26th, and Fox News now reports that Facebook hes begun its "death spiral". AllFacebook naturally disagrees, but I don't think the AllFacebois "get" the dynamics outlined above, reading their article.

The perfect Friday Lunchtime blogpost topic - white corpuscle sized computers in your bloodstream, courtesy of Ray Kurzweil and the BBC

Exponential growth in processing power and the shrinking of technology would see the development of microscopic computers, he said.
"We will see a billion-fold increase in the price-performance of computers in the next 25 years," he said.

and a few seconds later....

Mr Kurzweil said it was possible to accurately predict the growth and change in computing power by looking at how it had developed over the last 50 years. "There will be a 100,000-fold shrinking of computer technology over the next 25 years," he said.

Wow...accurate predictions of the industry, a 100,000 fold shrinking of the technology and a billion-fold increase in price/performance over the next 25 years - whats not to like!

Except.....I recall combining a laptop cum WiFi cum mobile device 10 years ago, and using Moore's law etc we confidently predicted the iPhone by 2004 for $250. Unfortunately industry structural barriers etc meant it didn't turn out that way - we got the iPaQ at $450m sans mobile instead.

Using Moore's Law's doubling every 18 months we can predict that the price performance in 25 years, and the likely size, will be c 1/200th of what it is now, and I can "get" that if we use devices that are really boned down we may get as far as maybe as much as another 1/100th shift - but b*gger me if I can find the logic for all those other thousands of times reductions.

It is my wont, when faced with stuff I don't believe, to translate it into LoLCat, as I find it makes more sense that way

"today u can put pea-sized computa inside ur brain... if u has parkinson disease n want 2 replace biological neuronz that wuz destroyed by disease."

he sed zilian-fold increase in computin performance n capabilitee ova next 25 year coupled wit 100...000 fold shrinkin... would lead 2 "blood cell-size devicez... that can go inside our bodiez n keep us healthy n inside our brain n expand our intelligence".

he sed blood cell computa would b able 2 "produce full immersion virtual realitee frm inside nervous system".

See...it makes perfect sense now

Lolcats iz in yr Matrix n pluggin into yr Brain............................

Umair Haque made a fairly interesting post earlier today, and this evening I thought I would comment, only to find he's removed it owing to blogsturm and drang - surely not, in such an august ecosystem;)

UPDATE - its on TechCrunch's Crunchnotes now, so here are my thoughts:

It was all due to the TechCrunch Effect, which Mike Butcher over at TechCrunch UK commented on - as Mike noted:

His argument is that TechCrunch is “peaking”, it doesn’t invest in the “community” and instead gets its attention by creating division. He calls this “The TechCrunch Effect”. Ouch.

Clearly the Bubblegeneration effect created some division too

In essence, Umair argues two points - Firstly, re that TechCrunch Effect:

I'm gonna make a prediction. TechCrunch (etc) are peaking. Without investing in the community - instead of just endlessly playing the community against itself - further growth (real growth, not just beta) is going to be more and more costly.

I'm gonna call this set of dynamics the TechCrunch Effect. It's the opposite of building a community. Instead of making a set of people with similar interests better off, you wedge them and divide them.

Actually, I'm not surprised there is a slowdown, growth at TC (et al too) has been phenomenal over the last 12 months - but competition is improving (blogs like this one and Umairs are both growing for example), and by definition growth from now on will be harder. I don't buy the divisiveness per se, its the TC schtick, and I prefer it to the parrot-the-news without comment of some of the others. They have an editorial policy and you knows what youse getting. TC has a community (just look at its comment board). What is probably more concerning is its role as startup kingmaker, but that too will pass.

The other point was re Etsy being the New Google:

If we're lucky, Etsy is gonna start emerging as the next Google. Microcommunities are going to explode. Etc. Why? Because at the edge, love is more powerful than hate - a lot more powerful.

I don't think a right-on Green Webshop is quite the same animal as a search engine - different parts of the ecosystem in my view - but I do agree that microcommunities will expand now, curiously enough because of Google's (and the New search players') ability to find them. Its a Coasian thing - as transaction costs reduce, it gets easier and easier to dive into the long tail.

Re the sturm und drang, Umair notes:

You get almost instantly tag-teamed by several people at once - guys who can spend all day blogging about you and your post, no less.

Well, I say carry on commenting - its always good to have thought provoking stuff to debate. And as Oscar Wilde said, there is only one thing worse than being talked about - take the link-lurve and run

Just into the Broadstuff RSS hopper is HipMojo's note that most Web 2.0 Ad based companies will fail, since - to quote:

...over time I became - via this blog - someone that others turn to to get a sense of where the ad market is headed.

In this capacity, I’ve also become an adviser to many startups, and occasionally, an investor that wants to test a hypothesis or conduct due diligence on a company they are vetting.

I must say: I am baffled with how much many VCs are neither realistic nor knowledgeable with the bulk of their advertising assumptions. Not only are the assumptions off, but they seem to fail to realize the psychology and dynamics that go into play with how ad sales work. The competitive nature and subjective method to the madness is generally unaccounted for. Or, alternatively, the challenges and complexities of building sales streams are wildly underestimated.

Just to add to this tail of woe, I'd add two more reductions in probability of success:

(i) There just is not enough Ad money to fund em all anyway. The Ad business is about $0.5 trillion globally, about 1% - $5bn - will go online globally per annum each year in the next few years (averaging out most forecasts). Most of that - c 90%+ - will go to the Big Boys in the game already, leaving about $500m tops for everyone with those $100m bizplans (and one with a $15b plan.....)

(ii) In any case, statistically at least 50% fail in 3 years, (up to about 2/3 of startups depending on who you believe).

It just gets better and better

Then of course there is the Founders Discount even if you do succeed......

And they say many people start their own companies because they can't stand the corporate grind - this tells you something about how dire corporates must be - but somehow I don't buy that they are that bad.

An interesting bit of research I've just been reading implies that optimism and risk taking may in fact be genetic - its possibly down to the D4-7 Allele:

There is now growing evidence that the behavioral traits which predispose some of us to risky and novelty-seeking behavior have a genetic basis. A recent book, American Mania, by a colleague, Peter Whybrow, director of UCLA’s Neuropsychiatric Institute, summarizes this evidence. He begins by noting that human migration is one major form of risky and novelty-seeking behavior. Only a few of our species left their ancestral home in the African savannahs and began that long walk to the ends of the earth which allowed homo sapiens to colonize the world. Who were these earliest migrants? It turns out they had a particular genetic profile. They had a higher percentage of an exploratory and novelty-seeking gene than those remaining behind. As novelty seeking and risk taking “are . . . behaviors essential to exploration and migration. . . this should be reflected in a distribution pattern of the relevant allele [the D4-7 allele gene] that is similar to the ancient migratory paths of our species.” How do we know this? The geneticist Luigi Luca Cavali-Sforza of Stanford University and his colleagues have provided a genetic mapping of the geographical dispersal of homo sapiens from their original home in Africa. Subsequently, Dr. Chauseng Chen of the University of California, Irvine, found that a coherent pattern emerges from this mapping “where those who stayed close to their original homeland have a higher percentage of the common D4-4 allele in the population and a lower prevalence of the exploratory and novelty-seeking D4-7 allele.”....

....This “migration” gene, as we may call it, is also found in those who evince the most extreme form of risk-taking and novelty-seeking behavior to which the gene predisposes its bearers: addictive behavior which often descends into manic depression (bipolarity).7 Risk taking and novelty seeking are of course also the hallmarks of the merchant and the entrepreneur. For both migrants and entrepreneurs are “mavericks. . . who run at the edge of the human herd. Migrants are a self-selected band of seekers—those of adventurous and curious mind—who in their restless approach to life lie at the extreme of the bell-shaped curve of behavioral distribution.”

So there you are....... now, if this is correct, then one way to foster entrepreneurial hubs is to set up conditions that attract D4-7 Allele holders.

I was reading Sigurd Rinde's blog today (we'd been chatting earlier about information flow - see his post on Dubious Information Distribution here) when I saw this post about sexiness vs profitability in software. Sig sez:

From my many years in business I learned one thing about this dichotomy:

Cool and sexy industries attracts a steady flow of new entries, players and investors. A kind of margin influx of money that is willing to be lost and people that are in it for the wrong reason. That makes it extremely hard to earn money, if anything at all.

Boring industries are the completely opposite, the people involved stays and know what they do, the influx of new entries and money is at most a trickle. Margins are stable.

So if you want to earn money, find a boring industry!

This can be drawn as follows (a curve I term the Reverse Cargill Position, in honour of Cargill Ventures, formed in 2000 by an agricultural goods company Cargill to invest in dot.coms)



However, I do think Sig is on less firm territory with respect to software when he notes that:

And now I'm in software. There you have two kinds:

Consumer software: The cool and sexy stuff, especially anything web based and 2.0'ish.

Enterprise software: The boring stuff nobody sees nor gets you anywhere when striking up a conversation at a cocktail party.

The implication being that Consumer = sexy = no margin, but Business software = unsexy = profitable.

What I can say for certain is that business software can be unprofitable as well as unsexy - some markets / competitive positions / business models are just crap, or after time become less competitive etc. Small new business software companies selling complex, non standardised software to large players is a typical example - in theory one can make money, but in my experience there are often 3-4 small players in any sector and in the shakeout process none of them make much money either.

Alternatively, if you can get those precious network effects with consumer software and achive that precious increasing returns economics, you can achieve market dominance and hit the G(oogle) spot.

However, Sig is probably right when he notes that, in order to achieve getting Fairly Laissed as well as Laissez Faire he:

....Would suggest "ski bum" to impress socially and enterprise software to impress financially. With one exception; if you're to see a VC next week, refocus your enterprise stuff to the consumer section and do not forget the colour scheme and Ajax while you're at it!

Maybe VRM is the exception, you can talk Social Networks and Death to Advertising while still exposing your CRM-like back end

Rory Cellan-Jones over at the BBC has an interview with Niklas Zennstrom about Joost.

Last year Joost raised $45 million from investors - after initial funding from Zennstrom and his Skype co-founder Janus Friis - to launch a platform enabling TV firms to put their content in front of a global audience. There was a lot of hype and hope about this advertising-supported platform which was going to be the big winner from the internet television revolution, just as Skype had made internet telephony take off.

But when I interviewed Niklas Zennstrom at a networking event in London this week, he conceded that Joost was going to take a lot longer to reach a big audience than Skype, partly because of the need to reach complex deals with media owners. And when I asked when he expected to turn a profit, he made it clear that was so far off it wasn't worth talking about now....

....Zennstrom insists that Joost does not need blockbuster programming, like the Premier League football on which BskyB built its business. He believes this global platform can prosper by serving viewers with all sorts of special interests, from fly-fishing enthusiasts to gadget fans. What's not clear to me is why a large audience will go to the bother of installing software without the lure of something they can't get anywhere else.

2007 was supposed to be the year that internet television took off, with the launch of a whole clutch of services with wacky names and big ambitions. But two things have happened - the viewers have been slower to take to online TV than expected, and the big established broadcasters have launched their own platforms, keen to avoid the mistakes of the music industry which handed the digital initiative to Apple's iTunes service.

Three things have stopped Joost so far according to our analysis:

(i) Their proposition has not got a good enough - or at least differentiated - proposition (or content) to drive switching behaviour, whereas Skype had an immediate benefit (Free)

(ii) There are a lot of new entrants by major playesr (eg iPlayer) - if anything, the "Skype" plays in Web TV are the newly emerging pure PC-TV plays.

(iii) The good enoughs are still good enough

For Joost to improve its fortunes, Content is always king - but UGC/ProAm is probably a better play than More TV We Can See Already.