We have looked at Enterprise  Social Networks in one of the earlier posts. While, there was feeling that enterprise networks mirror behaviour in public social networks, this is attempt to validate that premise.

Challenge about collecting data about public social networks (e.g. Twitter, Facebook, Whatsapp) is – all such networks are private with limited set of APIs. Whatsapp does not offer any API which itself is a challenge if you are looking to gain user insights by data analysis.

We eventually succeeded to extract data from Whatsapp, on much smaller scale though. This is really ‘back-of-the-envelope’ attempt to validate premise stated earlier.

Our sample size is to look at 15000+ messages circulated over six months in one of the Whatsapp group in this part of the world (India). This sample group is subscribed by about 40 members.

Considering 1 BN messages are circulated over Whatsapp in a day, this is really miniscule, so scope of analogy is limited. Though, would expect that at the individual group level, some correlation can be easily built.

  • Sample groups shares  100+ messages per day (including media) over six months. While there are some spikes which can be easily attributed to special occasions – festivals, sports events etc, there are no significant dips or gradual decline.

Users are engaged on day-to-day basis and contribute to the content of the group. While typical mobile apps would see spike in user activity and dies down over time if app does not add enough value every day of the user (think about how many times you really use banking app). Whatsapp seems to be buck that trend. With so many messaging platforms, it is most difficult thing to do.

Chart 1

  • Almost 15% of the messages are media types (photos, videos) as shared on the group. While text messaging still rules the roost, am expecting that media share would increase over time. It would be long way before users give up text messaging though.

Facebook report as shared in this post rues about lack of inclusiveness of the internet including local language content. We see that messaging happens predominantly in native language over Whatsapp. So, it is certainly speeding up inclusiveness by offering way to share local language content easily.


  • Let us look at the user behaviour of Whatsapp in this group. Taking cue from enterprise social networks, we can also look to slot users in below categories.

Observers – These are people of sidelines watching all the action and cheering in between.

Broadcasters – These are the people who write posts that tends not to attract many responses. 

Responders – These are the people who tends to  reply or like than post. If they post something, it is expected to gather less responses as well. 

Engagers – These are driving force for collaboration networks. They connect and sustain relationships. They balance ‘posts/replies/likes’ made with ‘replies/likes’ received. 

Catalysts – They energize the network by attracting responses. They receive a lot more than they give e.g. conversations that lead to lot of replies/likes. 

Being mature social network like Whatsapp, we should expect higher number of Engagers/catalysts than observers.

If you look at message count by each user in the group, almost half of the messages/content is contributed by around 7 users (engagers/catalysts), while rest is contributed by 20 users(broadcasters/responders). Around 13 users are not contributed at all , thus tagged as Observers. Percentage-wise, around 15-20% users are engagers/catalysts, 50% as broadcasters/responders, 30% as observers.


Almost 50% of the content is contributed by 15-20% of users, most engaged ones.

Chart 4

It seems natural progression of social media user where he moves from – Observer –> Broadcasters –> Responders –> Engagers –> Catalysts. Group becomes most engaged when there is healthy mix between these categories of users.

Finally, stickiness of Whatsapp is in its ease of use and very personal nature of messaging. In era  of information overload and un-solicited advertising over emails or other social networks, it will rule till it sticks to the core as Brian Acton (Whatsapp Cofounder) puts it simply,

“We don’t want to inundate users with messages they don’t want”.

That should be fairly easy thing to do.