TCS CodeVita : A quarter million pageviews later ...

TCS CodeVita Pageviews

10 days and a quarter million+ pageviews later, its time for analysis. Firstly, I think its a decent start for an initiative, open only to Batch of 2013 of engineering or equivalent graduates. The updated web page says the contest is open for only TCS accredited colleges. Assuming that approximately 4 Lakh students graduate every year and given TCS' track record of recruiting 30K-40K candidates for the past few years, I would assume that the target audience for this contest is approximately 2 Lakh students. The event is a team event where each team is comprised of 3 people. Given that there are approximately 6K teams, CodeVita seems to have attracted participation from as many as 18K students. Not bad, I would say. The registration deadlines have been extended to 31st July, 2012. Its very possible that CodeVita breaches the 5-figure mark. That's a huge number and I have not heard of any competition of that scale. It will be interesting to see how TCS manages to pull off things at such scale.

For the uninitiated, a coding platform usually comprises of an engine which compiles the user submitted code and runs a battery of test cases against it before it can figure out if the submission is to be accepted or rejected. To isolate users from each other, there is usually some sand-boxing that the engine does. All these activities - compilation, execution, sand-boxing etc. are resource intensive. Depending upon level of difficulty, its fair to assume that in a 6 hour window a team would submit at least 10 solutions. That will be 10 * 10K submissions for the engine to handle spread over a 24-hour window. That's roughly 1 submission / sec. To compile and execute 1 submission in 1 sec is a pretty stiff challenge. Moreover, the submitted code is written by the participant. It is very likely that these submissions would run longer. The longer a submission sits inside the sand-box, the more challenging it will keep getting for the engine to manage a rapid rate of submission. Will be interesting to see how its all managed.

Next, let me talk about region wise participation statistics. CodeVita website shows participation from each state. I have access to Google Analytics traffic report for this website. The next set of observations is based on data from these two sources and some correlation between them.

1. Bangalore and Chennai have nearly the same number of page views, but Tamil Nadu is far ahead of Karnataka in number of teams registered. What it probably means is that rest of the Tamil Nadu has done much better than rest of Karnataka.
2. Top 5 cities showing interest in CodeVita are Bangalore, Chennai, Kolkata, New Delhi and Pune while Top 5 states which have generated interest are Tamil Nadu, Andhra Pradesh, Maharashtra, Karnataka and West Bengal.  Hyderabad is sixth on the list. Not surprisingly, the states' growth is driven by their educational capitals.
3. That brings attention to Mumbai. Mumbai is a lowly 12th of list of cities showing interest. Excluding Navi Mumbai, Mumbai has about 25 engineering colleges in TCS' accredited list. We don't know the college-wise break down of registrations yet, but low traffic from Mumbai is a strong indicator of Mumbai's poor participation. I strongly suspect Mumbai's participation will be a tiny fraction of Maharashtra's participation. Given that TCS is head-quartered in Mumbai, this is significant. Wake up, Mumbai, you sleeping beauty ....
4. We see traffic coming from approximately 80 Indian cities. Would be fair to assume that each of those cities have on an average 3 engineering colleges. That suggests that we could see participation from approximately 250 colleges. In all likelihood, this could be the largest assembly for event of any kind where these many colleges have come together on the same stage.
5. 66% of the traffic is direct traffic i.e people have reached the site by typing the web site URL. 25% traffic is search traffic. 9% is referral traffic. Facebook has largest share in referral traffic. What this also means is that traditional means of advertising the event has been the pre-dominant form of advertising the event. 
6. Average time spent on the site is 15 min per user session. Not bad, considering that its only a registration site. This translates to 4500 eyeball-hours. (18000 users * 15 min)
7. Average page load time is a whopping 10 seconds. But this also includes photo uploads by registered candidates. Besides it also depends on end-user bandwidth. Many who accessed the site would also be first-time users. Their browser caches may also not be primed. Hence difficult to say whether the site is slow or not. My personal experience has been that the site loads in sub-seconds, because I have browsed it many many times.
8. 46% of the traffic is from Chrome, followed by 39% from Firefox and 11% from IE. In TCS (AFAIK), we don't use much of Chrome, so I doubt the development team was geared for this.
9. 95% traffic is from Windows machines, 3 % from Linux and 2% from mobile devices and corresponding OSes.
10. Web searches return hundreds of pages for "TCS CodeVita". Almost nil online presence for CodeVita on Twitter. 

Adios, Fellas. Will keep following and posting this event. If you like the idea of pan-India inter-collegiate coding competitions, please get the word out. Drive that budding geek you know to take part in the competition.

Disclaimer:- The author works at TCS. This blog is NOT an official source of information about TCS CodeVita. Since the author is a passionate programmer and the source of action is so near to him, it is but natural for the author to talk about it. Is that not why people blog?