Japanese twitter users at a football game. Source: theHindu.com (I see Ricky from "Better Off Dead" finally made it over to Tokyo)

I increasingly hear from people that they are “giving up” or abandoning twitter. Reasons given are the usual suspects: the signal-to-noise ratio is bad, no quality insights are possible in 140 characters, most tweets just look like shallow self promotion, product company spam. Indeed, there was a golden era (that lasted about 6 weeks) where one could tweet about a company, and receive a prompt response. Now that most companies have installed a twitter-monitoring robot, the service level is quickly degrading right down to the same level as the automatic email responders– or worse– the automated telephone system.

Twitter as a viable business tool is limited; as a marketing blurb tool it’s adequate, but as an ongoing small-talk medium it seems to do pretty well.

As I’ve picked up more and more Japanese friends on twitter, I’ve noticed that it seems to be doing relatively well in Japan (compared to the US market). Why is that? Is it just the stereotype of tech-friendly gadget-wacko Japanese? Or are there some other more specific factors in play? I have absolutely NO SCIENTIFIC RESEARCH to back this up, but here are my postulations. Perhaps someone out there can follow up with some actual useful information. I’m throwing this out now to get the conversation started.

Possible technology explanations for better twitter penetration into Japan:

  • Character/language density – One quality of East Asian languages that use Kanji is that many words can become pretty short. In English, “transport” is 9 characters; in Japanese it’s two: 交通. “Tokyo” is 5 characters in English, Japanese is two: 東京. Similar consolidations go for names, places, verbs, etc. Yes, Japanese grammar may get longer, but since when is proper grammar needed in twitter? It seems that the Japanese writing system has an inherent “density” advantage over English.
  • Input tools – Twitter seems tailor made for mobile devices: short, text-centric, quick input. For a networked society that already preferred the mobile phone-based Internet access, twitter drops right into place, often much handier than email. I’ve discussed this before.

Along with the tech explanations, however, there may be some cultural differences in play. These are kinda squishy, so please squint appropriately:

  • Inherent public posting – Most Japanese people I know guard their privacy pretty strongly, but no more so than Americans or Europeans, I think. However, there sometimes is a strong tendency to make idle chit-chat items public out to the group. Japanese people will usually not talk with strangers, unless someone breaks the ice with small talk about the weather or some ramen noodles. Then, the mode shifts to everyone wanting to build a connection with others, and everyone will pipe in their 2 cents. Mind you, this isn’t a political debate or anything concrete, just pichiku-pachiku. Twitter seems to facilitate this nicely. Let me put it this way: that signal-to-noise ratio that drives some of my American friends nuts fits in nicely with the low-level buzz that Japanese ping to each other to reinforce their relationships.
  • ongaeshi 恩返し — My wife (who is Japanese) dislikes instant chat with other Japanese, because of all the built-in politeness requires everyone to say “thank you” and “goodbye” at least 4 or 5 times before closing the conversation. For Japanese, a similar thing looks like it’s occuring in twitter, albeit in asynchronous slow motion: someone follows you, you actually have a bit of an obligation to follow them back. If someone tweets your name, you’ll RT a simple acknowledgement. Taken together, this may push enough people past that critical mass needed to keep a social network flourishing.
  • Lack of a Mixi / Facebook alternativeMixi is pretty noisy and cumbersome compared to Facebook, but has an interface completely native to Japan. Facebook, while cleaner, may be a little too “close” for the idle chit-chat that seems to go on in twitter (this is no different than the US: don’t you hate that guy on Facebook that has hotlinked all his tweets to show up in his friend feed?). Taken together, twitter gives a good outlet for quick updates via mobile phone in a text-rich environment (easily consumed on the train).
  • Son Masayoshi– The great thought leader of Softbank seems to be a prolific twitterer. Along with other leadingtech figures in Japan, Son seems to actually hold ongoing debates and policy discussions via twitter. This openness is refreshing if not counter-intuitive for Japanese business. Put together, twitter looks like the domain of the tech-forward progressives.

from the rightcliq website. What's up with the toolbelts?

Ever since Steve Jobs and Woz unveiled their little pet project to bring computing to the masses, we’ve had a debate about where the application should live: server-side or client-side.  IBM always argued that server-side is faster, cleaner, and generally more profitable for them.  Bill Gates made his billions bringing apps to the client-side.  Cloud Computing is really this argument wrapped in shiny new clothing– and has been blargged about ad nauseum everywhere else, so I won’t bother with that.

I mention that debate simply to bring up what might be a parallel argument that is forming now that many sites are becomming socially-aware.  Here’s the topic: is a social network invidivual-centric (client-side) or catalog-centric (server-side)? This isn’t a debate about where the actual software app resides, that’s pretty much invisible now.  For a social network, the core function is the graph– but where is that centered?  Is it centered around the individual (like facebook), or around the catalog (like amazon.com)?

An individual-centric graph has the person at the center, and she is free to add her friends, likes, dislikes, and catalog choices to her graph.  In this model, she will always want to log in with her personal network ID, and then interact or share with a given website.  Facebook is this way.  Pluck offers this kind of model.

A Catalog-centric graph has the products or new topics acting as the currency of the graph, and individuals may come and interact with them as they please.  The individuals may even use their Facebook ID for logging in, but the graph stays with the catalog and is structured in a way to build out the graph between those products, regardless of who or which individuals contributed any given part.  Bazaarvoice is a catalog-centric model

At first blush, it would seem that the individual-centric graph has run the table, but this may change. Faceboook offers a ‘like’ button just about anywhere (any major website) now, and people are certainly participating.  But those same individuals are starting to push back on the flood of information coming into their facebook feed.  Everyone is hungry for peer-driven invormation about products, but very few are willing to contribute content if it pollutes their friend feed.

Can a catalog-centric graph solve this concern for users?  Would you be willing to review more products or “like” more things if it could be somewhat anonymous?  What would you need in return?  Will you want to segment off your personal contacts (all your high-school drinking buddies) away from your shopping guru mentors?  Visa has announced Right Cliq, an individual addon that serves as the bridge between the individual graph and catalog graph.  Sure, we need another social network like another holein the head, but this may actually have some legs: consolidate your shopping peer-driven information with your purchase history, while segmenting it away from your personal contacts.

It becomes a catalog-centric social graph, but it belongs to Visa, not to the vendor.

p.s. if you like this article, please click the ‘like’ button in the upper right. :-P

Cocktail party by Mike Jones

When you’re at the pub, what do you talk about?  Your car? Your weekend?  Your golf game?  Your new tasty favorite indie band?  FSM help you if you talk about work– boring.  When you’re online, what do you chat/write/blog/tweet about?  Sure, we geeks talk about the biz and tools and sites and Steve Jobs gossip– but that’s part of our job.  If it really were free time, what would you really write about?  And where would you write it?

Watching the online online communities mature, I’ve noticed that people progress along a path:

  • First, they blabber anything just for the sake of blabbering.  This is really just experimenting with the toolset for most.
  • Second, they share everything they see and read and link, until they realize that everyone else is linking/sharing the exact same stuff.
  • Third, people settle into their “thing”: constant updates of their children, their home business, political rants, mindless blogospam.  Here’s where people’s idiosyncrasies start to show up.  To quote John Worfin, “Character is what you are in the dark.”
  • Eventually, people resort to the same topics they like in the pub: politics, their children’s sports, luxury vacations, and making up historical  facts.

That covers Facebook and twitter for most people (twitter is more gossip-y, but whatever).  But what about an online community based on a catalog ecommerce site, or a specifically-themed site?  The one thing all the visitors to that site is the stuff that’s for sale, so this becomes the dominant topic for the online community.  At some sites, this works, other sites bomb miserably.  Why?  What is the difference?  Where is that line?

Likely:

  • Woot.com – wins because all visitors are bargain hunters, and are happy to share juicy details on the deals.  We see this in real life where people brag about the coupons they clipped or the steals they got at Try-n-Save.
  • SteapandCheap.com – wins because of the same reasons that woot.com wins, but even better because these people all share a leisure activity: they ski (or snowboard or hike… er… they go outdoors)
  • Any catalog that is basically a technical sale: consumer electronics, exercise equipment, software, new cars, gardening tools.  If you think you’re the neighborhood expert, congratulations on becoming the new King of the Online Community.  Online dashboards are great for this.
  • Politico.com – wins because all those ranters who alienated their friends on facebook with their hatred of Presi-senato-gressman Blankenstien now have a home where they can win Free Internets with the other crazies (just like the bars in Dupont Circle)  Online dashboards are also great for this, but only to show who is the biggest d-bag.

Not Likely:

  • VictoriasSecret.com – Who wants to brag about their lingerie?  TMI.  Questions & Answers generally wouldn’t work for personal apparel.  The most obvious question “will this fit me?” has the most obvious answer: “I have no idea who you are, so how the hell would I know?”
  • Vehix.com – mostly used cars, which means that deals are not repeatable– it does no use bragging about your good deal, no one cares. The same goes for any one-off item.
  • fandango.com – tickets to movies (which people _do_ discuss at length in pubs), but Fandango is a ticket site– nobody has seen these movies yet, and they certainly don’t want to hear the ending.  BTW, Inception is all inside Leo DeCaprio’s head because he’s insane!!

So, will your website be able to build a community successfully?  Will you really be able to plug it back in to increased user-generated content about your products?  I would suggest that the ‘tell’ is rather simple: would people talk about your goods in a pub?  Would they mention your stuff at a cocktail party?

truer than you think

Times was, I used to be able to spout off any random factiod I thought I knew, and the Internet took it as read truth.  The Wikipedia used to be great for this.  Now, my rants are pretty much limited to the blarg you’re reading right now– we’re probably all better off for it.  Most university professors scowl very deeply if a student references the Wikipedia in a footnote, which is fair, but not for the reasons most people think: Wikipedia is a bad reference source because it’s a derivative work, not because it may be inaccurate:  The student should be citing the original work, not someone’s summarized boilerplate.  Wikipedia has largely squashed the ‘inaccurate’ label through a zealous use and requirement of all statements must have footnotes.

But that raises a conundrum for many of us: where is truth?  Where is the expert?  Is the expert the one with the most experience?  Is the expert the one with the most money?  The most devotees?  Is truth simply the mob’s consensus?  Graduate school told me that truth is the logical sum of a tested thesis.  I spent 15 years being smug that I knew what that meant, only now to really see that when the Internet gave everyone a soapbox from which to preach, now social networks are giving everyone a Hyde’s Park corner complete with audiences.  Companies like bazaarvoice and pluck are setting up these cacophonies wherever possible (good for them).  These systems invariably include meta-rating systems to rate the reviews and the reviewers, in the hopes of crowdsourcing the good information from the bad.  In general, it usually works.  It is still, however, all based on a Kuhn-model of mob truth.

The NYT recently published an article on a new computer named “Watson” designed by IBM to play Jeopardy.  Another possible use they summized might be to find counter-factual statements to anyone’s gtiven declaration on the Internet.  In short: a bullshit detector.  I can imagine they will be able to monetize this thing into millions of dollars: every social network and review thread can now come with a robot that can read plain speech, offer immediate counter-responses to erroneous information, and perhaps even show us a numerical score for ‘trustability’ or ‘truthiness(all the footnotes in that link– irony!). The downside here, of course, is that most reviews for most products will be reduced to little more than the barren subjectivism of American Bandstand: “It’s got a good beat, I can dance to it.”

There is still salvation for quality content in quality reviews: hard numerical data, solid logic, and qualitative feature polarization.  I’ll explain myself on those in some upcoming posts.

© 2010 Dave Jenkins contact me via twitter @davejenk1ns or via email blog at davejenkins dot com Suffusion WordPress theme by Sayontan Sinha