Tried it with Instagram instead of Facebook. Also pretty interesting. https://www.google.com/trends/explore#q=%2Fm%2F05pdgbz%2C%20...

I did the same thing. It sort of takes one's breath away.

What I find interesting is that Twitter and Facebook have the same graph shape, both peaking in 2013.

They should allow a log scale, then you could compare with Facebook more easily.

My fairly crude theory is that those sites have slightly broader and thus less tech-savvy userbases, so searches might represent ingrained ways of accessing the site more than they would for Reddit or Tumblr.

As even the least-tech-savvy becomes more aware of faster ways of accessing the websites, and browser software starts to pre-empt user intention more and more, this kind of convoluted search starts to drop off.

Not sure if that theory holds water, or is in any way necessary.

I agree with your theory to an extent, however:

> As even the least-tech-savvy becomes more aware of faster ways of accessing the websites

I'm not sure that happens. From observation, people tend to stick to the first way they figured out how to do something on their computer. Searching for "Facebook" from Google is very standard practice. Further, browsers _encourage_ search over direct address input. I've tried to show people many times how to visit sites directly... it doesn't help them so they understandably ignore the advice.

Also, Google stopped indexing Twitter around the time Twitter started to decline, but Facebook still comes up first for many name searches.

It's the impact of mobile.