Couple of weeks ago I met a team of entertainment industry people in Bangladesh who are working on to replicate their success in the offline world online. We talked a little bit about their content strategy, and I felt that this startup is going in the wrong direction. But I also knew, if I said this on that moment then these industry veterans will rule me out immediately – I simply don’t have enough experience about the entertainment industry to give such opinion. So I just listened and moved on to another topic.
But the issue kept bugging me for the next few days. My feeling about the issue was, not just these guys but the whole TV industry in Bangladesh is doing it wrong. Importance of some types of contents is highly overrated. People are spending huge money to make such content which Bengali viewers don’t want to watch. But who am I to talk this ‘nonsense’ in front of people with years of experience?
So I decided to do a simple analysis on Bangla (by Bangla I mean Bangladeshi content only; at this point the title might seem to be a bit misleading, but I have my reasons behind it) content viewership and interaction on YouTube. The analysis is targeted for all people interested on real user data. I understand that because of Bangladesh’s internet usage trend and the nature of the data, Google’s finding that ‘web traffic trends online represent the offline world’ probably doesn’t apply in this case. Added to these, we have not considered data from other video sharing services. Nonetheless, I believe, these data are better representations of Bengali viewers’ tastes, and far more reliable compared to the current TRP measurement techniques in use locally.
Now let’s dive into the data part!
I have collected YouTube video data in 8 categories from the first ~450 items in each category using the video streaming service’s API and a simple Python script. The result was a sample size of 3553 videos in 8 different categories – Bangla movie, Bangla music, Bangla TV drama, Bangla crime shows, Bangla ‘talk show’, Bangla cartoon, Bangla news and Bangla magazine program.
The analysis was pretty straight forward. I just wanted to know two things.
· The market-share for different types of Bangla content on YouTube.
· Which type(s) of content generate highest amount of interactions?
But how to calculate market-share for different content types? And how to define user interaction? For simplicity, I used view counts as the measure of market-share, and comments as the measure of user interaction.
The results —
- The total view counts of the 3,553 videos in my sample set exceed 193 million.
- Bangla movies have the highest market-share – 30.5% (58 million+ views), followed by Bangla music with 21.1% market-share.
- However, Bangla music content has the highest interaction – 30% of all.
- Crime show content should be followed closely – it secures 16.6% of total views.
- Talk shows and magazine programs have the lowest market shares with only 1.6% and 2.5% of total views, respectively. (In my opinion, TV channels should probably rethink their content strategy!)
You can view the numbers in simple pie charts below. If you have any comments or queries about the data and/or analysis feel free to share it with me at firstname.lastname@example.org.
And here is the full summary of the stats: