Alex, ---------------------------------------------------------- which of the following four offers the biggest market? which potentially will offer the highest profit margin? ----------------------------------------------------------
Consumer applications probably offer the biggest market for what you're trying to do but also present the highest risk for you as a small business in need of cashflow. If you can solve an Enterprise's problem, they will basically pay your product development costs (altho these can appear as "consulting fees" on the invoice - just be careful about any IP clauses in the contract). Most enterprise software has 1 or 2 "sugar daddies" in its development stage - and you can often see the influence of these sugar daddies in the products many years later. The thing is: You have to be focused on solving their problems rather than doing the cool stuff that you want to do. The question I have, which I am not wholly clear on - what differentiates your technology from other text analytics tools (and that's everyone from Autonomy to Leximancer to Radian6)? What do you do that other people can't? Looking at the 2 areas that I know most about: - HR: to understand the informal networks that rule the company outside of the formal hierarchy. Only a few HR departments are advanced enough to be interested in this. There are already tools on the market (such as Trampoline which is resold by Customware here) and these are struggling to get traction. - Knowledge management: to find the people that have the 'soft knowledge': There are a lot of tools that are in this space from very different backgrounds (enterprise search, ERP, ECM, social bookmarking, SNA, etc). Most tools do a mix of things - e.g. Automony is primarily a search engine with fancy semantic analysis capabilities built on top. It aims to help manage 'hard knowledge' and do the tacit stuff as a bit of a freebie. Tools in this space struggle because the market (senior managers) have a poor understanding of how knowledge-based work actually happens and hence little interest in spending money on it. Cheers, Matt On Wed, Jan 6, 2010 at 7:59 AM, Alex Dong <alex.d...@gmail.com> wrote: > @Phil Sim, @Joz Ong, @Matt Moore: Thanks for sharing your thoughts. > >> What your really looking to pull out is some kind of authority or influence >> and frequency of keywords in someone's stream is probably not a very good >> indicator of that. > I definitely agree. > This reminds me what happened to Google vs. Altavista war. Basically > Google > won because they figured out PageRank is a better way to raise > signal/noise ratio. > What you're suggesting here, especially the authority rank by retweets > and > potentially the number of followers, are as important to our engine as > PageRank to Google. > >> I think if you want to demonstrate the power of your engine, and I assume >> you can do this, you need to work across social networks IMO. If I can put >> in a term and it gives me results across twitter, facebook, blogs, linked in >> and just as importantly shows me how the influencer uses the different >> networks in respect to this term, its starting to bring some real value > I can't agree more. One more point here: statistics works better when > there are sufficient data. Large number rules. > > > This has been said, however, if we're targeting enterprise customers, > for example, law firms, to mine their internal communications, I'd say > that > today's twendly is good enough. It works pretty well when searching > for a long tail expertise in a relatively clean signal environment. > Whereas pulling data from blogosphere and other content source will > significantly increase engineering cost by almost an order of > magnitude. > > I'm glad that this discussion has advanced one step further: > Given the four markets below, assuming we could deliver for both > cross social platforms and enterprise level service. > ---------------------------------------------------------- > which of the following four offers the biggest market? > which potentially will offer the highest profit margin? > ---------------------------------------------------------- >> - Marketers and PR: So they can get into the Word of Mouth circles >> - HR: to understand the informal networks that rule the company outside of >> the formal hierarchy >> - Journalists: to find additional leads or people that can get them to it >> - Knowledge management: to find the people that have the 'soft knowledge' > > Cheers, > Alex > > > On Jan 5, 10:43 pm, Phil Sim <philip...@gmail.com> wrote: >> Interesting that you cite autonomy as being a different type of play. They >> actually have an Expert Locator in their product set, so I'd suggest they'll >> be a major if not your biggest competitor. >> >> As Twendly stands now, I don't see a great deal of value add as opposed to >> just doing a simple key word search on Twitter and pulling the names out. >> What your really looking to pull out is some kind of authority or influence >> and frequency of keywords in someone's stream is probably not a very good >> indicator of that. >> >> I think if you want to demonstrate the power of your engine, and I assume >> you can do this, you need to work across social networks IMO. If I can put >> in a term and it gives me results across twitter, facebook, blogs, linked in >> and just as importantly shows me how the influencer uses the different >> networks in respect to this term, its starting to bring some real value add >> to the table IMO. >> >> However, I think you really need to take it to the next level and start to >> bring an authority measure into it. So if I tweet about Silicon Beach and >> then that gets many retweets, then I've got more influence than someone who >> may tweet more frequently about the topic but whose network is small, or >> whose network doesn't recognise me as being as influential. Now if you can >> do this across networks, taking into account blogs/links, facebook/likes, >> etc. then I think you've got a pretty compelling offering. >> >> My application actually represents one of the use cases Elias suggested and >> we're moving in this direction (at the moment planning to use opencalais to >> decipher meaning) so feel free to give me a bell if you want to chat >> further. >> >> >> >> On Tue, Jan 5, 2010 at 2:54 PM, Alex Dong <alex.d...@gmail.com> wrote: >> > Great conversation. Thought I might be able to provide some background >> > of what we're doing. >> >> > Let's forget about twitter for the moment. I'd prefer to leave it to >> > historians to make a judgement on whether twitter is providing value >> > or not. As Tim has mentioned, we're using twitter as a data source so >> > that we could tune our engine and demonstrate the engine's potential >> > via 'real, live, massive' amount of data. We are not married to >> > twitter and within weeks, we could deliver a Twendly for blog, Twendly >> > for Confluence, Twendly for exchange so on and so forth. What's >> > important here is what the underlying engine can do. Data is only a >> > skin deep. >> >> > Let's also forget about consumer vs. enterprise play for now. Windows >> > has home edition as well as site licenses. Quicken has QucickBook and >> > Quicken as well. I firmly believe that the problem we're solving >> > exists on both side of the firewall. It's just a matter of given our >> > experiences and resources which direction we're best positioned to >> > attack first. >> >> > Let's focus on what's the problem we're trying to solve: >> >> > +++++++++++++++++++++++++++++++++++++++++++ >> > We are helping people to find each other. >> > +++++++++++++++++++++++++++++++++++++++++++ >> >> > #1. We're not a "information search engine" like Google or Autonomy. >> > We find people instead of content. Of course we'll provide a preview >> > of the content for one search result, but it's only for 'prove to me' >> > purpose. >> > For example, if you search forhttp://twendly.com/?q=siliconbeach+ >> > "lifeguard+paper", >> > you will find 5 people who are talking about "Lifeguard paper" at >> > almost the same time. If you click into any of them, you could see >> > what they were talking about. >> > People first, timeline, historical info are all secondary. ( Tim's >> > famous "People Centric Software" idea. ) >> >> > #2. We're not a 'real time people search engine'. We believe in the >> > value of aggregated historical data. >> > A good example here is a search for "Tiger Woods":http://twendly.com/?q= >> > "tiger+woods" >> > Now if you are looking for someone who tweets more golf than gossip, >> > you'll definitely give #5 a try since the rests are obviously >> > interested more in celebrity than sports. >> >> > #3. We're not a 'content aggregator'. We might need to bring in >> > contents from multiple sources, call it 'break down the asymmetric >> > info wall", or 'bridge the silos' or 'identity fragmentation', that's >> > only our crawler's mission, not ours. :0) Ours is to apply a "people >> > search" on top of all the bits and try to make sense of it. >> >> > #4. Our current goal is to help people to find "influencers" and >> > "connectors" within a community. ( The Tipping Point, anyone? ) The >> > whole internet could be one community, this google group could be one, >> > a 300-employee company or a 150,000-people corporation all can be >> > treated as a community. As long as there are people, there are needs >> > to find each other, to connect. >> >> > Having said all these, the engine still needs significant improvements >> > on both accuracy and scalability. We're still trying to find our >> > first beachhead. Please give twendly.com 5 minutes, play around with >> > it and share with us your answer to this question: >> >> > --------------------------------------------------------- >> > What does Twendly.com enable you to do? >> > --------------------------------------------------------- >> >> > Cheers, >> > Alex >> >> > On Jan 5, 12:27 pm, "Hendro Wijaya" <hendro_wij...@hotmail.com> wrote: >> > > Hey Tim, >> >> > > This is just my thought and I might be wrong. >> >> > > 1. Twitter is half-fad. >> > > I can't see how Twitter will grow through the same trajectory like >> > Facebook. >> > > They won't die, but, won't be as mainstream and active in the consumer >> > > space. >> > > The most sustainable use cases I see people using Twitter are: business >> > > self-interest (tweet their blog post, deals, etc) and once-in-a-blue-moon >> > > real-time news. >> > > Effect on Twendly: sustainable value proposition. Why would I, as >> > consumer, >> > > keep using it? I only use Twitter Search Engine when there is an >> > earthquake >> > > (or some real-time news that I care). >> >> > > 2. Tech-acquisition as the only way out. >> > > Continue from 1. Without sustainable value proposition, I think it will >> > be >> > > hard to monetize and leave you with tech-acquisition as the only way out. >> > > For me, that is a very risky bet unless you think you guys can build >> > > something way better than Google, Bing and Twitter in this space (e.g. >> > > Etherpad's case). >> >> > > 3. Solve real problem. >> > > I personally made this mistake and it's been a valuable lesson. >> > > Based on my own experience and your video about HiveMind, I think >> > > asymmetric-knowledge is a real problem in enterprise. I think it will >> > worth >> > > more in long run if you guys focus on solving that. >> >> > > Again, I might be absolutely clueless and please feel free to enlighten >> > me. >> >> > > -Hendro >> >> > > -------------------------------------------------- >> > > From: "Tim Bull" <tbull...@googlemail.com> >> > > Sent: Tuesday, January 05, 2010 8:27 AM >> > > To: "Silicon Beach Australia" <silicon-beach-australia@googlegroups.com> >> > > Subject: [SiliconBeach] Long time SB member, new entrepreneur seeking >> > your >> > > assitance on ideas for our site >> >> > > > Hi All, >> >> > > > Thought I'd re-introduce myself - I've been a long time lurker in this >> > > > community (I was one of the early sign ups when Elias launched it). >> >> > > > Recently, I quite my job of 15 years at PwC and co-founded >> > > >http://www.binaryplex.comwithAlex Dong - we are focussed on the >> > > > problem of expertise location. Initially we thought this would >> > > > primarily be an Enterprise play and have collected some interest in >> > > > this space along the way, but we are wondering now if there might be a >> > > > consumer play here instead. >> >> > > > We've created a new Twitter search enginehttp://twendly.comwhichis >> > > > indexing Twitter users and using our underlying people search engine >> > > > to demonstrate the concept. >> >> > > > I'd really appreciate anyone here who could spare a few minutes to >> > > > have a quick look and a poke around at it and let us know your >> > > > thoughts either through the feedback link on the site, or e-mail me at >> > > > tim.b...@binaryplex.com. >> >> > > > To give you a taste, a search for silicon beach >> > > >http://twendly.com/?q=silicon+beach >> > > > reveals these are the people who've tweeted most about it over the >> > > > past 12 months. If you view this on the site you'll see both a >> > > > relevancy rank and a histogram. >> >> > > > #1 Mick Liubinskas @liubinskas | Australia Web business builder for >> > > > Pollenizer - global, early stage tech, services company. >> >> > > > #2 Pieter Peach @ppeach | Melbourne, Australia Learning to build >> > > > things. Doctor intrigued by the web & decision making. Founder, >> > > > SimpleSponsorship.com. >> >> > > > #3 Mike Casey @mikecasey | T: -33.923693,151.19054 Director at >> > > > GradConnection, Tech startup fan, HR tourist, internet analytics >> > > > machine. Yet another R1b1b2a1a2f >> >> > > > #4 Ian C @aussie_ian | Sydney I love travel, the web, and music! >> >> > > > #5 Daniel Purchas @danpurchas | iPhone: -33.893589,151.146988 >> > > > Director of a Graduate Recruitment Online Marketing Company >> >> > > > #6 John Haining @johnhaining | Sydney, Australia Dad, Business >> > > > Owner, Entrepreneur, Dashing Geek :) >> >> > > > #7 ryancross @ryancross | Sydney, Australia a geek, duh! >> >> > > > #8 Tom Voirol @voirol | Sydney Online strategy, social media and >> > > > user experience consultant. Husband, MBA student, kitty litter box >> > > > cleaner. >> >> > > > #9 IPitch Australia @IPitchAU | Sydney, Australia Where Australian >> > > > entrepreneurs and startups come to pitch ideas, find investment and >> > > > get the latest industry news >> >> > > > #10 socialmediameme @socialmediameme | Social Media Strategies >> >> > > > Would appreciate any feedback you could give us and suggestions on >> > > > which direction we should take this idea - our gut instinct is that >> > > > there IS a play here in the consumer space, but your thoughts and >> > > > feedback would really help. >> >> > > > Cheers, >> >> > > > Tim >> >> > > > -- >> > > > You received this message because you are subscribed to the Silicon >> > Beach... >> >> read more » > > -- > You received this message because you are subscribed to the Silicon Beach > Australia mailing list. > > Guidelines on discussion: http://tr.im/ujKF > > No lurkers! It is expected that you introduce yourself: http://tr.im/ujMm > > To post to this group, send email to > silicon-beach-australia@googlegroups.com > To unsubscribe from this group, send email to > silicon-beach-australia+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/silicon-beach-australia?hl=en?hl=en > -- Matt Moore M. +61 (0) 423 784 504 m...@innotecture.com.au
-- You received this message because you are subscribed to the Silicon Beach Australia mailing list. Guidelines on discussion: http://tr.im/ujKF No lurkers! It is expected that you introduce yourself: http://tr.im/ujMm To post to this group, send email to silicon-beach-australia@googlegroups.com To unsubscribe from this group, send email to silicon-beach-australia+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/silicon-beach-australia?hl=en?hl=en