It really is a bit hard to not watch the Prisoner’s Dilemma cant look away Twitter Car Crash whose Feb 12-June 9, 2009 “rough meeting notes” Techcrunch are collaboratively leaking with a gun to the “Nervous (not alert) System’s” 1_Billion_user_target (valued at $1 each) : “that will be the pulse of the planet” SouthParkCo. While it’s nice, in regards to staff, that they want to “Make sure people are happy” it is not 100% that Engineers are at the top of Twitter People Pyramid in the way they are at Google from Founders down. (so said a quote i cant find as it’s not a searchable post!)
Reading thru the longform depositionish Eric Schonfeld post with morning milked espresso + creamed honey toast, I was struck not by the textbook Acquisition targets that have been to fill product gaps (CoTweet for enterprise multiple account login / Twitpic for images / Tweetie for iPhone) - Nor was it the BigCo + Diddy Celebrity partnerships + entertainment advisory positions envisioned or that “How do we communicate to the Board (and investors) to back off” was actually written down. Further, while the acquisition/IPO triggering event meeting points are interesting (they want to keep operational control post event), but what I found most interesting is the centrality and unsolved solution around the classical search problem of Relevance vs Recency which then gives rise to a myriad of product, mission, business model, architecture questions.
It seems Twitter is searching excuse the pun deep into itself.. both at the twitter.com consumer destination + the Commercial API “every tech company is interested in Hosebird” enterprise search + content syndication levels to solve the relevance vs recency issue. Let’s face it most of Twitter search if not all of it is straight boolean search - Pretty much Lucene out of the box - which is becoming more of an issue as the spammers, sextailers, emarketers, get_new_follower_pyramid_bots and all other version of not quite human terminators minus the ability to identify and kill John Connoresque humans are making the product worse (Anyway it seems Twitter is rightly concerned that google with its sizeable search leadership (like apple did by porting OSX to wireless handsets) will do realtime relevance search better than the once Summize small team of PHDs once they have the same 100% of all tweets with no API or latency restrictions- Whether they will give Google a leg up in return for a Google/MySpace or Microsoft/Facebook type 8 or 9 figure revenue deal.. is an interestingtrade-off fo’sure.)
Good vs New : - “Google would kick our ass at finding the good tweet.” : Notion of a “Good” search result - indicating a trade-off between relevance and recency, as well as the authority/insight of the source which is a key objective to offer by year end “Tweet rank.” (possibly similar to the vgood Twinfluence ?) Not without merit, Twitter positioning wise see themselves as “what is happening right now” versus “Google is old news.” Although they recognise “Google has Realtime Content (stocks, weather)”
(GOOG, MSFT, YHOO) Partner vs (Twitter.com) Destination : “Playing with fire here where we know that Google is building the competitive product.” - Does Twitter build out its consumer marketing to get people to Twitter.com - rather than have their content syndicated to the largest websites in the world - Google, Yahoo + Microsoft; “They want to put this on Google.com” Ironically, Twitter is facing a newspapers type dilemma - Receive the traffic and revenue from google, but maybe be neutered by them in the process. But they also dont want to follow MySpace and become irrelevant which being closed is the first step in becoming.
Twitter vs All Microblogs eg Facebook + Friendfeed : “Would facebook be included (?)” + “They are thinking about relevance as more than just chronological.” “Fusion of different types of content, relevancy is key.” “There will (be) some weight on recency.” - Fascinating question at a Corporate Strategy Level of whether the Winner, if there can be such a thing in The Real Time Web - will it be a One Trick Pony such as Twitter that does one thing extremely extremely well - Twitter fulfilling it’s Nervous System Discovery + Sharing vision -
Or will it be the Hosebird enabled, Facebook Connect, Disqus, FriendFeed SUP, gnip piped player (namely google on the search/discovery side and a much more open statusy updated wireless beefed up Facebook possibly with a friendfeedish acquisition on the microblogging side) that is able to suck in all the sources of realtime content with no (noticeable to the second) latency or delay, and then provide a standalone recency (twitter only) filter or an integrated relevancy weighted Social Media realtime search result - that pulls status updates from Facebook, Twitter, Blogs (posts and comments) plus whatever viral video, forums, flickrs are flying around - With the punchline being a contextual self service monetisation adsense like system ?
It’s like a Seesmic.com case study at the multi-billion dollar opportunity level : Seesmic was a realtime video P2P consumer destination, but with a strong ability to raise capital it was able to aquire a leading Twitter centric third party app that also increasingly centrally pulled in your content from your multiple distributed social networking identities/profiles. So does Twitter follow a Seesmic multiple content types centralised model… or keep its 140 character group text message focus - The Techcrunch Twitter documentation makes it appear the team believes pulling in the various related feeds its members have (similar to how friendfeed does it I’m assuming they mean) is a commodity offering. However, it’s #1 Fear Factor coming from the 310 documents is when Google (and Facebook secondarily) receive the Hosebird pipe of realtime comprehensive Twitter content : That when this is bolted into + emulated/extended into the Google Real Time Social Media Offering.
Discovery (as opposed to “Distraction” which they mention alot when referred to Diddy, marissa Mayer + Microsoft hehe - what a combination of bus dev stakeholders) There was one interesting set of bulletpoints about business vs engineering decision making and its negative affect on “happiness” for which a committee was suggested at being formed. Nuthin like a happiness committee.
“Discovery and sharing goes hand in hand.”
“Discovery of tweets and Discovery of People are both Important”
“Discovery of Tweets is more intriguing.. Relevance is the first problem.”
“How do you propogate any tweet or message to the Right Person” - The Hunch which comes thru as The Big Capability that Twitter want to build into it’s offering + shift the cards in their favour is Discovery.
Typically this would mean suggested content; suggested followers; suggested hashtags; and suggested ads the Dave Winer hope of ads which arent ads but business solutions to your consumer problems. The nirvana of a relevant search result before u enter the keyword.
“Maybe Machine Learning, saved search.”