LinkedIn big picture with Python

Have you ever tried to analyze your LinkedIn profile and have a bird’s eye view on all your LinkedIn connections? Couple of month ago, I started wondering about who my contacts were and how I could classify them and get the overall idea about my own LinkedIn profile.

The irony is that more contacts we have -less we know about them. LinkedIn provides a vast amount of detailed information, but at the same time it doesn’t allow to see a bigger picture. That is why I decided to solve this problem by creating a tool which analyzes all my connections and presents it as statistical information.

This is how GetSocialNetwork project was born. The project is Open Sourced and is written in Python. It provides a framework to gather and structure publicly available information. Once I ran the program I was surprised to find out so many interesting things about my own profile.

When I opened my LinkedIn connections list, my first question was “Where are these people are coming from?”. That was where I started.

country stats

The program outputs a simple txt file which I bring in Excel to visualize the data. I didn’t expect to see UK contacts outnumber my Italian contacts but what surprised me most was that I almost didn’t have any contacts from CIS countries! That was quite unexpected considering that I worked with so many people from there.

The next thing I wanted to know was where my contacts were working.

company stats

As I work in media industry most of my contacts are in VFX and animation companies. I was truly surprised to find out that I have 22 connections from MPC. Wow, I guess MPC is so big that it is inevitable to stumble on someone from there.

Next thing I wanted to know was the titles of my contacts.

title stats

Ok cool. A lone scientist on the bottom of the list made me wonder who it was. Later, I found out that he was a data scientist. Probably, he would laugh his head off looking at my homegrown data methods.

Next thing. How much experience do people have in my contacts? I have almost 10 years of experience behind my back that means that the majority of my contacts should have at least the same amount of years. This is what my app showed me.

years of exp

And finally, I wanted to know how often people switched their jobs. I decided to monitor the period from 2005 to 2015.

work switch stats

I have no idea if it is good or bad that there is a very subtle decline of people switching their jobs. Looks like with age the life gets more stable (not my case sadly). Or maybe, there are less opportunities to find new jobs?

Here is how the program works.

program pipeline

The has additional recursive methods to obtain contacts of contacts of contacts and so on. The thing grows exponentially and arguably may grab the whole LinkedIn database. However, it will never happen because LinkedIn has routines to prevent people from mining their network. I tested the methods and they work and it takes time to gather multiple levels of connections, but I do not recommend to overuse these methods!

It was a fun project to work on and I learned many interesting things about my profile’s trends and stats. The project is still rough and buggy and it only supports English language, but I leave it as it is and move on to the next adventure.

Leave a Reply

The reply will be processed within a day