One of the themes in The Disruption Dilemma, is that when a firm faces disruption, it is unlikely that the Clay Christensen promoted solution of setting up an independent business unit to compete with startups and itself in that new space will work. There are many reasons for this including, most importantly, that the last thing an incumbent wants is to speed up competitive pressures against itself. Another reason, as Tim Harford recently documented, is that adjustment can be hard for incumbents.
Enter Facebook. It tried something new. Faced with potential disruption from Instagram and later Whatsapp, it acquired each. However, rather than integrating them immediately, it kept the CEOs and Founders on and attempted to operate them as independent units inside Facebook. But for some shared infrastructure they would be pretty much unrelated to Facebook proper. For Instagram, in particular, this proved to be very successful going from 30 million to some half a billion users in 6 years. Moreover, part of that shared infrastructure was Facebook’s mobile ad solution that provided Instagram with the monetisation it was otherwise lacking.
As of today, the founders of Instagram and Whatsapp have left their companies and Facebook. Autonomy could only last so long and eventually those companies had to be brought into the Facebook masterplan. This was to be expected. Even companies that have tried their own independent business units (such as IBM) have had to bring them back inside in a painful integration. It is not clear how painful that integration will be in this case, but one thing for sure was that Facebook was not immune to the ‘more often than not’ phenomenon of acquired founders leaving some short time after acquisition. In this case, 6 years is surprisingly long.
I suspect this is all part of Facebook growing up and moving out of the startup phase where it believed that it might do things differently and learning that some things are hard to change. It will be interesting to watch the changes to see precisely what issues led to this particular integration.
Digitopoly 
Sounds like Facebook’s approach worked very well, from your description of it, both in terms of user numbers and monetization. Maybe there is just a natural progeession, as the company “disrupts” its original model while learning enough from the independent business unit to be prepared to integrate it?