Usability, User Experience and Social Web

Twitter is turning to a gigantic echo chamber (and that pisses me off)

Sat, 14 Aug 2010 00:36:00 +0200

I don’t really know what game they’re playing at Twitter. But it looks like they’re trying to turn a service able to share information at the speed of light and create unlikely conversation into the kind of sterile echo chamber that would make MSN and the whole makeup blogosphere jealous.

Everything started when they decided that tweets starting with @johndoe would only be visible by people who already followed John Doe.

I do understand tha what I’m saying to John Doe may not interest other people. But seriously, if I wanted to talk with John Doe privately, I would use private messagess, email, or any other damn private way to contact him. Public messages are supposed to be seen by everyone, and that’s how conversation start. Blocking my replies to people who don’t follow the user I’m talking to just make conversation a little bit poorer. Maybe this comes from technical constraints. When Twitter started to filter replies, they were constantly down, but anyway, Twitter was way better before.

More recently Twitter has started to display a list of users I should follow. Great, really: suggested people are people my other friends are already following. That’s how social media should actually work.

Too bad every people I should follow – but French minister Nathalie Kosciusko-Morizet – are people I already know in real life and I have chosen not to follow.

I don’t know which of them started first. But I’m quite pissed off because I can’t definitely drop suggestions I don’t want to hear about. Everytime I’m back browsing the list, I always meet the same list of people. Soooo boring.

By preventing me to discover new people, Twitter is definitely turning to a gigantic echo chamber. Don’t take it personnaly if you’re on the list of people I chose not to follow. I have a very narrow attention, and can’t follow more than 150 people at a time. I consider Twitter as a daily watch tool that generates conversation. I don’t want it to turn into another IRC like just because I can’t follow anyone but people the politburo wants me to follow.

Original article writen by frederic and published on Usability, User Experience and Social Web | direct link to this article | If you are reading this article elsewhere than Usability, User Experience and Social Web, it has been illegally reproduced and without proper authorization.

4 good reasons to run a private beta

Wed, 11 Aug 2010 13:17:00 +0200

Is launching yous product with a private beta still relevant in 2010? I was discussing this point last night with a real beta sceptic. He said a private beta is an old fashion way of generate some buzz, and make people talk about you for free, or almost. Sooooo 2004, and overhyped since Gmail invitations eventually ended on eBay.

I, indeed, disagree. Running a private beta, even a short one, is the most clever thing you’ll do in all your launch process, and a critical one unless you’ve got lots of cash to invest pre launch.

Why is it so?

1. Don’t run a private beta for the buzz

Forget about it. Unless you’re Jeff Bezos, you date the whole Techcrunch staff, or your product is a real game changer, it’s impossible for you to get a massive amount of free advertisement through a sole private beta.

Private beta Gmail style lies on a very old and simple principle: by restricting the number of invitations, Google creates scarcity. What’s rare is expensive – and we know how rare a cheap horse is, therefore how expensive it is. Scarcity generates envy. A Google product can do it, a new free online music streaming can do it too. But forget about it, you’ll probably just fail at this.

What you can do, if you manage your invitations the good way, and if your service is good, is having people talk about you – therefor generate some interest – as well as some quality backlinks, which is always good.

2. Test your scalability

Opening your product to a small amount of people at a time allows you to manage your scalability. Whether they come from your product or yout hosting platform doesn’t matter. Do you remember France.fr? This is definitely the kind of site that needed a private beta to test scalability but could not because of its nature. Doing load tests before releasing was not supposed to be an option either, but that’s not my point.

Last year, my old friend Bastien and I started Twitmark, a social bookmarking service that was doing the link between Twitter and Delicious / Diigo. We had like 30 people beta testing the service, and we quickly realized it could not scale. We managed an incredible amount of data, and our adtabase schema was not good. We’ve been forced to rewrite everything from scratch, and finally stopped the service because we lacked time.

Our beta testers showed us our limits, but without suffering from them. That’s what was important.

3. To manage feedback and motivate your users

For ages now, software publishers thought their clients could play paying beta testers. A beta is not a free QA process, and the word “beta” should not mean “unfinished buggy crap”. During your private beta, testers should only meet edgy bugs, and give you feedback about what you should improve regarding your product behavior or feature. Remember: your developpers may run the best tests ever, they will always be different from real users testing.

Private beta allows you to drive users feedback in an experimental environment. Controlling how many testers you are makes you controle the feedback flow.

A private beta also turns your users into beta testers. You can ask beta testers a much more detailed feedback you could not ask regular users, and come back to them to get more feedback after delivering your changes.

4. To validate your model

My final point. Your beta testers allow you to validate your global model before releasing it in the wild.

When I started coding with SCM, the first thing I was told was commit early, commit often. When I launched my first open source project, I was told to release early, release often.

I recently met a 21 years old serial starter, who was running his fifth company. Having sold 3 of them, and failed once, he told me: fail early, fail often. He had closed his third company a few hours after launching his site on private beta. First users showed him that his global model, which looked great, was actually a total failure.

That’s the final purpose of a private beta: check if your model’s validity.

Win 20 exclusive invitation to Play Nicely, the project management with fun Web application

Mon, 09 Aug 2010 20:32:00 +0200

Last month, I did a review about Play Nicely, an online tool introducing a splash of fun in project management. Still in private beta, Play Nicely just announced a new release coming with fancy feature, such as Github / Subversion integration and achievement badges for the best bug squashers.

I’m thrilled to announce you that Play Nicely team just gave me 20 invitation code to give to my fellow readers, to test their new version. To win one of those, you just need to answer the following question:

Which of those job did I never do?

Warehouseman in an alcohol depot
Archivist at an insurance broker
Bartender in a gay club
Worker in a printing factory
Barista at Quick
Subscribtion seller at OFUP (French company selling magazine subscribtion with a huge rebate, definitely a student job)
Dog keeper for depressive dogs

Don’t try to send my resume to Google translate, I didn’t mention any of them.

Is Foursquare already dead? Aka the Sick Sad World of useless social networks

Sat, 03 Jul 2010 20:25:00 +0200

Is Foursquare already doomed? A few days after the geolocation service announced a 20 million dollars fund raising, the question is already relevant. Not because of Foursquare itself, but for every social network having its 15 minutes of fame before falling into the pit of oblivion. A few weeks ago, my Twitter timeline was full of my contacts check ins, mayorships and badges. Today, it is as silent and gloomy as a battlefield after a war.

Let’s analyse why Foursquare – and a majority of social networks – are doomed, not for a lack of business plan but by design.

Foursquare model is the same as many social networks. It’s a well known one that anyone can reproduce ad nauseam.

To be successful, a social network mus attract a critical mass of users. Once reached, this critical mass will naturally attracts new users until the market runs dry: you’re opening a Facebook account because your friends are on Facebook. The social network starts attracting innovators, then early adopters, early majority, late majority, and finally the laggards. Reaching the mass market implies its growth curve will finally slow down at some point. Nothing new here, it’s just Roger Innovation Adoption Curve.

To attract innovators, Foursquare use 3 well known levers:

The innovative lever. We’re talking about geolocation, and more precisely about iPhone geolocation.

A community lever. Community is the very base of a social network. Foursquare is brillant as it first aimed at iPhone owners, which can be considered as a social class in the Marxist way. iste du terme.

An achievement lever. Foursquare relies on both competition and achievement. Competition between users to become the mayor of popular places. Achievement because your reputation growths as you’re winning badges: for checking in 5 Apple Store, 50 different location…

The 2 first levers are used to attract new users. The third one is the most important since it’s used to ensure their loyalty. Once again, there’s nothing new here, but Foursquare shows its mastery on this particular topic. Unfortunately, this well known model has 2 major issues.

First, because it considers that the innovative concept it relies on will reach the majority market, driving it to the majority market at the same time. In theory, this is a win / win principle of mutual tension: if Foursquare drives geolocation to the mass market, then the mass market will associate geolocation to Foursquare. It usually fails, but works great sometimes. Apple’s Ipod has a very strong link with MP3 technology ; iPod is almost a common noun when you talk about a mobile MP3 device, even though it wasn’t the first to hit the market. Apple just made a wonderful use of the above mentioned levers:

First iPod relied on an innovative technology: MP3, which was supposed to replace CD but had not reached the majority yet.
iPod created a great community feeling: early iPod adopters used to recognize themselves in the street to the white earphones.
iPod allowed its owner to take all their music everywhere.

Second, 95% of them are both useless and their own purpose, so easily subject to fatigue an disinterest.

3 years ago, when I wrote about Twitter just being 2.0 masturbation (in French), I missed a critical point: Twitter is a medium, not a purpose. Twitter is just a way to communicate adapted to a market and an era, like IM, email or avian carrier.

Foursquare is its own purpose. For its users, Foursquare only purpose is to check in, become mayors, kick other Foursquare users and play the bigger badge collection contest. Playing with Foursquare is easy and cheap: you just need to use your cellphone and tap 3 times. But what’s important is the fact that you need to check in.

2 years ago I was discussing with Bertrand Duperrin about how difficult 2.0 adoption in enterprise was, beyond technological issues. He mentioned this fundamental principle of social media: people contributing to social networks are people having too much time on their hands. Hence the difficulty of enterprise social media adoption outside of a defined business project, but that’s not my point. However Foursquare users have exactly the same profile: they need time. We travel a lot to go from one point to another, but we don’t really care of where we pass in between. We use travel time to do something useful, usually not travel related, unless we find something related and useful to do.

Foursquare interest is very limited. Foursquare interest in a regular use is even more limited and cat rapidly turn to be boring. That’s a critical problem for emerging social networks, starting with Foursquare: how can they retain innovators and early adopters long enough to get the critical mass of users necessary to be noticed by the majority? This is the only way for social networks to apply a business model usually based on advertisement and affiliation.

That’s the problem Foursquare will have to solve – as many other innovative products – if they want to survive beyond their investors cash.

Why doing simple and cheap when complex products sell more expensive?

Sat, 26 Jun 2010 15:39:00 +0200

Being an advocate of lightness and simplicity sometimes makes you forget about the “hard market reality” (sic). A few weeks ago I was heaving a lunch with the CEO of a software company who explained me why complexity was his core business:

We designed a complicated software on purpose. It gathers every feature our clients may dream about, and we add every enhancement they ask for. But deploying it needs days of a certified consultant. Everybody think we’re selling software. That’s wrong, we’re selling certification and consultancy.

It’s sooo simple that the hardest was to think about it.

Play Nicely: if coding is fun, managing code should be fun too

Sat, 26 Jun 2010 15:09:00 +0200

In 5 years of managing software development, I’ve tested almost every bug tracker and project management tools on the market. Most of them are either unfinished craps or unusable gas factories. Despite a huge choice available, my experience has been quite disappointing so far.

Released in private beta 2 weeks ago, Play Nicely is yet another hosted code management tools claiming that Coding is fun. Managing it should be too. Play Nicely is brought to you by Adam Charnock, Rob Hudson and my friend Basheera Khan. Play Nicely relies on limited feature, a nice user experience and a fancy interface to rejuvenate both project management and bug tracking.

I met Bash during LeWeb 09, as she was working as a freelancer journalist for Techcrunch UK. We eventually became friends after I used my evil Briton bait: a good old iron teapot filled with Earl Grey ;-). When Bash changed her carrer path and told me about Play Nicely, I made her promise to give me an invitation away. I’m happy she kept her promise as it shows that her project is now live.

Play Nicely’s statement is interesting, it is also questionable. Coding is fun as long as it’s just adding new feature. Things get more complicated when it comes to fix bugs and reread old code. However, I must say I love huge bulldozer driven refactoring processes as long as I have a good test coverage to help me. So offering a nice and intuitive project management tool should be a way to make the hardest and more boring step easier.

Once you got your Play Nicely invitation, creating your account is trivial: just fill in your first name, last name, a username and a password – no confirmation – to access your account. Play Nicely does not support authenticating with Facebook / Twitter / any Oauth provider yet which is really strange talking about a geek oriented product in 2010. I hope this point is on their roadmap. Even stranger when you realize that you can’t upload your photo yet but are forced to use Gravatar instead. The email I’m using on Play Nicely is not linked to my Gravatar account so I’m now stuck with the horrible avatar. That sucks.

You won’t understand the user interface at first sight, but the way the elements are placed is very clear and gives the feeling that everything is where it should be. Some elements should be redesigned to make the interface more fluid, but the general impression is very nice, which is exactly what it’s supposed to be.

Getting started is made easy thanks to a very well done 7 steps tutorial. Play Nicely invites you to discover it by managing a fake project tasks. 2 minutes after registering you’re able to manage your first project without asking yourself any question.

Play Nicely is mostly divided into 2 universes: the Playground gathers both roadmaps and tasks from your current project on a single screen the Doris way, and a lifestream shows everything happening on your projects.

I’m not fan of life streamings on ticketing systems, even though they offer filters. I’ve been using such a lifestream on Jira, and logging every change on a task makes it impossible to read. I do thing adding lifestreams to project management tools is more a fashion issue than aiming at efficiency. In my opinion, a simple table with every task status is really enough.

Let’s talk about the tasks as they’re the core part of Play Nicely. I’ve been really disappointed not to be able to create a task without assigning it to an existing milestone. I’ve been forced to create a bulk milestone just to write down things I wanted to add my project later. On the UI side, creating tasks is trivial and you can reorder them or move them from a milestone to another with a simple drag and drop. Nothing extraordinary in 2010 but it’s still simple stupid as I like it.

Tasks management is both the best and the worst part of Play Nicely. The interface sucks, and I had a hard time finding how it worked the first time despite a very light form. Adding a description or a comment should be on 2 different screens, and the task details should be at the bottom of the screen.

This is really bad as task management has also its cool side. I’m definitely a fan of the railway like status choice, as I love being able to make internal links to other tasks just by mentioning “task X”. Making an item a task or a bug is a 1 click thing, and both have their own color to differentiate them. Play Nicely is definitely good for these small details.

So, will I leave my good old bug tracker and switch to Play Nicely? No. First because Play Nicely still can’t plug to my code repository which is a fondamental feature. Their website say Play Nicely supports SVN and Git, I didn’t find where or how. But most important, because the interface is too fun, and lacks the “solid as a rock” side I’m expecting for a project management tool.

Facetime is coming, are we ready for mobile videochat?

Sun, 13 Jun 2010 15:14:00 +0200

Video chat is an old science fiction fan dream. It has been existing on cartoons drawing what would be year 2000 since the end of the 19th century. Being able to chat face to face making fun of the distance may be an old dream, it has a hard time becoming true despite not technological issue.

Despite the fun and ubiquitous aspects, we don’t like videochat. Operator efforts in 2005 to make it the new hype after France Telecom (main Frech telco company) attempts to make it something popular at the end of the 90’s eventually resulted in a big fail. We don’t like videochat because it reveals what we want to hide. It shows that we’re not dressed, that our make up is a mess, what we think about the other speaker, where we are (and shouldn’t be), that we have our finger in the nose, or totally absent.

The times, they have changed. We’re now used to shout where we are and what we’re doing, with who, and how we’re feeling, hoping for a massive amount of people to know about it. Facebook, Twitter and Foursquare have changed how we define privacy. But is it enough to change our mind about videochat, even more videochat on a mobile device?

Steve Jobs seems to think so. Adding videochat on iPhone 4 is an audacious move form Apple. Cupertino has always been what people wanted before they knew it, but I think they’re now one step ahead. There’s a line between innovation and revolution that Apple only crossed twice, and it was a failure. Videochat on iPhone is not a tech revolution, it’s trying to revolution the way we consider a cellphone. I’m dubtfull about this.

In Stephen Baxter’s Transcendant, videochat is not about you coming with your environment, but about you being integrated into your contact environment. Videochat helps to mind the cultural gap, showing you next to your contact place, speaking his language, and dressed the way people dress on his planet. Maybe that’s the solution to combine both videochat and privacy.

I didn’t want to write about iPhone 4, too many people did it way better than I would do, but Facetime changed my mind as it leaves too many questions to be answered. What do you think about it?

About Web performance and user experience

Mon, 07 Jun 2010 18:20:00 +0200

Web performances has been a trending topic for the past 3 years, both as a user experience and a SEO matter. Many domains started to raise awareness on a slow Web being a critical issue, as people get used to get their content instantly.

It took a very long time for Web standards and XHTML + CSS to become a development standard VS. table layout + inline style. On the contrary, Web performance has grown to a major issue in a few months only, since they had a real business cost. According to Forrester Research survey, a page loading in more than 3 seconds means 40% more people stopping the checkout process.

How did this happen?

Forget about real time Web, instant information and all the marketing bullshit you’ve heard about for 2 years. Things are a bit different here.

Performance awareness comes mostly from the Web becoming a mass media, and integrating itself to our daily habbits.

The general accepted the Web being slow as long as it was only a mere source of information or a handy replacement for things we used to do another way. The Web was considered as superfluous and somewhat experimental. Today, Internet has been integrated into our most current habbits: press, shopping, communication, becoming an essential part of our lives. Companies are now relying on the Web as a critical part of their infrastructure, replacing desktop software with Web application. What was superfluous is now a standard, and slowness is not bearable anymore.

Available bandwidth on domestic lines has incredibly increased – at least in France and Western Europe. Standard bandwidth raised from 512Kb/s to or 20Mb/s, making slowness a real pain on the users point of view, but also increasing global page weight.

E-commerce revolutionized mail order selling, suppressing the boring mail part. Even though you could order by phone, the operator was still a an intermediary between your purchase and you. Web suppressed this intermediary, making things even faster. E-commerce sites had to be faster as well.

Web performance in user experience takes various forms:

Page generation and download time.
General elements download time.
The way browser behaves when manipulating elements on a Web page

The time taken to generate a page and download the various files can be easily tracked. These measures are indeed relative to the bandwidth available between the client and server. Measuring download for a 56k modem can seem useless in 2010, even though…

There are many tools vailable to track loading times, starting with Firebug. Firebug network tab gives very interesting information on how the various elements of a page are loaded. Firebug itsel has various plugins dedicated to Web performances, each doing measures according to a set of best practices : Yahoo! Yslow and Google Page Speed. You can also test your Website against both of them using Gmetrix.

Yslow and Page Speed best practices should not be taken as gospels: they are only general guidelines and do not apply to every site and application. Some can even had the opposite of desired effects. Setting up a CDN can be desastruous if it is too far from the application users, with shitty transit lines.

Google Webmaster tools has integrated an interesting performance chart since page speed is considered in Google algorythm. This chart comes with precious advice on optimizing significant pages.

This chart measures the whole page loading time, including external elements such as pages included in a iframe. Here, you can see both times when I added Facebook Like and Open Dislike buttons. Both of them are loaded in an iframe, and, even though users don’t feel it, they are part of the whole page generation.

Server side compenents loading time can be precisely measured, but the way a user feels an application behavior is something definitely subjective. A fast application having a heavy interface will seem slower than a simple one. Users want both rich application with video and fancy Javascript effects and client side performances. Old browsers will have a hard time managing rich contents like Javascript and DOM interactions. That’s the reason why the glitter part must serve the user experience, not be here for itself.

Once again, you can learn best practices from e-commerce Web sites, starting with Amazon.com. Amazon uses some fancy effects – mostly dropdown menu and a carousel – but they are optimized and give an incredible feeling of speed, even on older browsers. The result is: page scroll is never slown down. Effects are not on a page because they’re fancy, but because they have a real business value.

Mobile Web explosion was a major backward in terms of client resources and available bandwidth. This was expected, but the consequences were not. Instead of creating fast, optimized mobile versions, software publishers built traditionnal application for the main mobile operating systems: Android, Blackberry, Symbian, iPhone and Windows Mobile.

This is not really a surprise, and it may be the best solution against the lack of resources. Applications embed all the interface components, and only need to load the page content on the mobile device which makes loading way faster. This also takes into account the fact that users are not really used to using a browser on a mobile device while they’ve been using applications for ages. However, these applications usability and performances are ofte poor, which makes the user experience all but satisfactory.

What eBay iPhone app says about mobile shopping

Sun, 06 Jun 2010 09:52:00 +0200

eBay is the most popular mobile shopping in the world, having more than 8 million downloads on the App Store and counting. After traditional bids and classified announces, the biggest auction website now attacks the mobile shopping market.

In addressing this new market, eBay faces 2 challenges. The first one is to give its power users efficient mobile management tools. The second one is to please user who did not use the Web version.

Power user will be fully satisfied. eBay app is probably the best eBay experience ever. You can follow your auctions and place your bids while not being in front of your computer. The main part of the application id a complete dashboard giving access to everything you need, starting with critical information.

The application design is very well thought, with the dashboard elements also being quick access buttons. Being able to access your favorite searches directly is also a nice feature.

As a critical feature, the search engine is present everywhere on the application. It is fully configurable, even though its setting is a bit complexe. Searches are done on both auctions and announces by default, which is not eBay Web default, and I think that setting is a bit strange.

eBay is using iPhone push feature to send real time notifications about bids evolutions, allowing users to place new bids. I just lost a very nice lens 30 seconds before the end of the auction because I did not catch the 3G network in the metro fast enough. So everything on this app was made to have eBay users spend their “offline” time on the site.

eBay on iPhone also targets potential users the site was unable to attract in its traditional shape. The real time following on bids is quite addictive, making them funny. eBay turns a very serious business into game, giving the feeling you’re playing at a virtual casino, with a touch of excitement.

The product datasheet is much clearer and readable than its Web counterpart, and make us forget the Xmas tree style of certain eBay shops. The application gives access to high resolution pictures and product description, which is what you can expect from such an application, but that’s not the main point.

The most interessant point is that big blue “Place your bid” button, which makes purchasing fun and exciting. You’ll click, won’t you? Clicking on that damn button is really tempting. However, eBay plays it responsible by placing it under the product information, under the screen waterline. The button placement is really well thought, making the bid a 2 times action:

I must read the product information.
I can buy

Then come information about the vendor, shipping and payment. This is important since it shows a deep change in eBay way of selling since its first days.

Since its creation, eBay has been relying on karma, which is eBay’s buyers and sellers reputation. eBay’s karma indicates the eBayer’s reliability showing the notes and appreciation he had for every transaction on the site. On eBay classic, this information is in a widget placed at the top of the page, on the right, where 65% of the users have naturally their mouse cursor. This is the most important place of a Web page. On eBay for iPhone, this information is at the bottom of the page, under the product description and calls to action.

This is a major strategic turn from eBay, explained by the fact we’re on a mobile application. eBay philosophy on a mobile platform is radically different from what it is on the Web platform. On eBay classic, the main principles are trust, security and completeness of information. eBay defines mobile commerce with fun, ubiquity and real time.

Is Twitter trying to kill the ecosystem that made it successful?

Mon, 24 May 2010 16:29:00 +0200

As a young platform in a constant (r)evolution, the Web is a very interesting environment to observe. There is no need to wait for ages to see history repeating, mostly when it comes to doing mistakes. The day Microsoft included Internet Explorer as a core component of Windows, it killed innovation in the browser area during 5 years, unless Mozilla eventualy came with a mature and not too geeky product. There were, here and there, some attempts to counter Internet Explorer hegemony, starting with Opera, but their market share and impact on the public where anecdotic.

Twitter making Bit.ly its default URL shortener was a strategic move for the first one, and an attempt to survive for the latter. Bit.ly was far from being the most popular URL shortener, but it had the best analytics available. With 60% tweets coming from Twitter Web interface, Bit.ly quickly became your link tracking best friend, without killing the competition, if you except Tr.im. 40% of Twitter users are still using third party clients using their own URL shortener or letting users change according to their needs. That’s the reason why alternate URL shorteners survived Bit.ly acquisition.

Being quite busy lately, I missed 3 importants information affecting Twitter ecosystem. The first one is Twitter acquiring Tweetie, a very popular Twitter client for iPhone in April. The second one is Twitter releasing an official Android application. The third one is Twitter releasing an official iPhone client, based on Tweetie 5 days ago.

Official reasons behind Twitter for iPhone launch are interesting: a real demand for an official client and the poor user experience of third party Twitter clients.

Careful analysis of the Twitter user experience in the iTunes AppStore revealed massive room for improvement.

Comprehensive analysis of the Twitter user experience in the iTunes App Store showed very plainly that people were looking for an app from Twitter — we didn’t have one so they generally got confused and gave up.

Deciding to control your clients user experience from A to Z is the common strategy at Apple, with iPhone third party applications forbiden to compete with iPhone base applications. Furthermore, this is not the first acquisition in the small world of Twitter clients, but it’s the most important one because Twitter changed the rules of the game.

As a company, Twitter has always been handicapped by its very nature. Twitter is not a valuable service by itself, it’s “just” a notification platform. People send notification other people can subscribe to and relay. Twitter success comes from a gigantic ecosystem of third party applications. Everything started with a Web client allowing users to send public and private notifications, and receiver being able to get them via SMS. That’s the reason of the 140 characters limit.

Then, Twitter API allowed the blossoming of an incredible amount of third party services and applications, from mobile and desktop clients to silly games and useless services – therefor essential to give Twitter some fun. A simple Web interface and a gazillon of mobile clients enabled a huge community of passionate users. This led to conversation, eventually making Twitter the new conversation platform. Despite the noise, signal was strong enough to make Twitter a both technologic and marketing watch platform for brands that accepted to listen what people were saying about them (I don’t talk about so called influencers and all the conversationalist marketing bullshit; Greenpeace VS Nestlé on Facebook would prove anyone that a good social media campaign won’t make a mega corporation stocks drown. Did you stop eating Kit Kat? I did, but only because I’m on a diet).

Twitter crossed a line by releasing official clients. The company is now competing with its ecosystem, which should look slightly unfair, but perfectly understandable (or maybe not as I’ll explain later). Unless Twitter would close its API – and commit suicide – that move shouldn’t affect existing users. This moves aims at futur users, and finding new users in the war against Facebook, if there’s a war. I first thought Twitter would rely on Tweetie being a paying client as a way to monetize the service, but they made it free as soon as they bought it, which makes me wonder what’s their true strategy if any.

It’s hard to understand why Twitter is messing with the ecosystem that made its success. Making some application official will just kill innovation. Not selling them won’t make them earn money. Furthermore, releasing an official client 2 days after Seesmic – which I find way better by the way – was a stupid bet. Either Twitter would stifle Loic’s announcement or be unnoticed. I heard about Twitter client from my RSS feed as no one in my timeline even mentionned it. I’m now waiting for Twitter to show some coherence in buying its own derivative products, as it’s not clear at all for now.