For example, the biggest problem is with their “auto-logging capabilities” on third-party sites. To illustrate,   i happened to go to CNN.com and Yelp.com today - and i was really surprised to see that both the sites knew who i was and showed my friends on facebook who like that page or were on that site. Now that is fine but how did CNN know who i was? It happened to be that i was logged into Facebook.com at the same time. Facebook is sharing login session information with sites that are implementing its social widgets. But hey, i didnt say yelp.com or cnn.com should know my identify just because i am logged into FB. Tomorrow any random site can implement the social API and if i happen to visit that site, it will know who i am!

In fact this gets even worse!. If Facebook was just showing my identifiable information on the third party site through an iFrame or something which didnt let the third-party really have my data, it would still be okay. But third-parties are actually getting your profile information directly from FB. See example of Yelp above (i have smudged some of the profile info ..). If you see the javascript files they load, you will notice (see the blue highlighted text) that Yelp is actually getting all the data about me from FB without having any permission granted to access that data.

Third-party sites should not be given default Opt-IN by Facebook! And what is i happened to press the like button by mistake and that action gets transmitted to my entire friend network in an instant. Every technology should offer ways for users to correct any mistakes - there is non - once you press “like”, its gone to the activity black hole and you cant even retract. This has huge implications which only different users mishaps will illustrate in time

I am excited about the whole concept in general but the privacy issues are quite a barrier for now. Infact now i am going to be doubly careful to notice any facebook “like” widgets on sites - make sure i dont press anything by mistake - this is definitely going to make my experience weird to say the least.

A better way would be to have the ability to “tag” files and folders just as we tag so many social media items on the web. It would be great, if one could just right click a file or folder and add tags to describe its content and then be able to search files through a tag search. Such an application, lets call it FileTag could be very useful in making document organization simpler and scalable.

Update:

Read on the Techcrunch about the Tag Application for the Mac that does something like the imaginary FileTag. Time to go checkout it.

A usual strategy is to keep developing in a small team, adding several features, because you want to get the best thing out. However, this is usually a never-ending process. One can get sucked into a development/brainstorming/refining cycle and this usually happens because there is no real users yet who are using the product - that is there is little possibility of organic growth within the product itself based on what users like and do.

Then arises the second question? How does one launch - try to create a sudden big bang or go for slow but surely rising one can and then one day when you are ready, email the top tech bloggers like TechCrunch, GigaOm and hope that they cover you. Very likely you will see a traffic spike if any of these bloggers write about you but that spike can be a drastic fall too (see the graph of cuil.com as an example).

Just got a link from a colleague about How to Launch Your Software? The blog described the strategy adopted by Gmail developers to launch their product. I dont necessarily think there is one better model - in some cases a big bang may work but i am more inclined to follow the slow but rising traffic model that is based on the key principle - “get your users from day one” . That is a key point and i have seen the benefits of this in my own several startup projects. Users are the acid test for your product and will surprise you by not using or recognizing the importance of a feature that you thought was “cool”.

For example, we built this very cool flex based interface to show the connections between people and people and food for ifood.tv - we called it the ifood explorer.

iFood Explorer

But we soon realized that people didnt care much for something cool - they just wanted the recipes first and then anything else.We have not involved users till completing the product and so learnt a good lesson, of getting users as early on as possible.

Ok, the logo is for an institute for oriental studies. The idea of an oriental institute with the red sun does convey the purpose of the company - an institute of learning of the orient. But boy, the logo conveys a totally different perspective, if you see the sun as an enormous backside, and the institute spire, poking it (other other ‘p’ word, would sound too nasty). What do you think? This logo has been discussed elsewhere too

Being a startup guy for 3 years now, one of the most important thing that i have learnt is that, simple rules, that may sound as cliche, are actually very very important. There are quite a few of these simple rules or rather mantras, which i will discuss as i go on, but today time to reflect on the importance of the name and the logo. A sticky, meaningful logo can convey a great deal. A good logo should communicate the core idea or mission of the company. And in my experience, just hiring a designer (specially some of those who offer to do your logo for say $50) wont cut it.

Infact, before you even hire a designer the startup founders should be very clear about who they are and what they are offering. Many times, startups fail to define their core mission and goals and get bogged down in daily grind of “Getting things done”. However if the mission and goal is solid, everything flows from it, even the idea for the logo.

One golden rule of know what you are really doing is to write it down in just 2 lines (cut it down to 1 if possible). Once you have this clear, you could generate visual ideas and communicate the same to the designer. But just telling a designer we are to come up with a logo doesnt cut it.

Here is an example if a logo for one of my startups, ifood.tv .

When we made this logo, we really werent sure what we were offering. The logo doesnt communicate the core idea of the website at all - it just tells you the name. Farming this logo out to a designer would just have made it prettier, but not more informative.

iFood New Logo

Our new logo communicates the central idea of a site that is video based and allows any one to create their own food network. The first iteration was done by my friend (is if referred to on this blog as COP) and then on along with the inputs from the team and an outsourced designer, it  got its new final shap.

The new logo is better than what we have, but still has a long way to go. Some designers are really particular about the company clearly defining who they are. And i like that. A friend , who is a top brand and product development guy, a heck of a creative guy, really nailed down what ifood.tv is (or ought to be ) in terms of its communication message (will add the logo he built) . However, that design never got completed, because it was too clean, i guess.

Here are some of the most communicative  logos, i have come across

FedEx has a great logo - Notice the “->” arrow embedded in the letters E and X. Clearly communicates the speed at which fedex delivers.

The logo of Sun microsystems is tied to their tagline - to the Nth level. logo has every letter built of the letter “n”. Will post more.

The cricket world cup was up in 2003 and then unless you bought some
expensive dish network subscription there were hardly any options to
watch the games. Online videos were non existent.  And even if one
could watch the game at some friends house, the time zone difference
between USA and cricket playing nations was such that one would have to
watch the game are really odd hours (i remember the world cup games
used to start something like at 3 am). So i always used to wonder,
wudnt it be just easier if one could watch a 30 minute highlight of a
cricket game (the one-day versions are 8 hrs long) online?

Solution Goals:

1. To develop a website, www.crickethighlights.com (i still have the domain) to showcase the game highlights

2. To develop a software  solution that would automatically extract
highlight segments from the video broadcast. This would require fair
bit of video processing but wudnt be terribly difficult. One could
check the signal for the score and determine points where a player gets
out or hits a four or a six (just think of them to be like home runs in
baseball).

Status:

1. I didnt do much till 2005. then found a digital video processing
guy who was doing a PhD in the area of video summarization. Sounded
like a perfect guy to build the technology. After many mths of trying,
we cudnt really get the software going. Mainly because he got busy with
his PhD.

2. 2006. Pitched crickethighlights.com as an online video site for a
b-school course on launching new ventures. The idea was very well
received but since this was not cricket and not baseball, the feedback
was that the market is too small.

3. Nothing since then. In 2007 the BCCI, announced it would develope
a portal for doing exactly this for the global cricket audience. Seemed
like this was the end of the road for crickethighlights. However, the
BCCI site never really took off

4. then Youtube happened and suddenly one could see 2-3 minute clips of the cricket games.

So perhaps, crickethighlights.com is now an idea that should retire to the deadpool. What do you think.

UPDATE On May 15, 2010: I have been getting more than 200 or so odd unique visitors on this domain, i am embedding a widget to help awareness of new diabetes treatment trials

You can find more such trials at TriaX.com