A new kind of #FlashHacks event: Mapping corporate networks

The #FlashHacks event in London on is going to be different than the other ones you’ve attended. We’ve noticed a lot of you are very interested in mapping corporate networks. However, because we usually have a coding strand and a data hunting strand, the corporate network mapping gets left behind. This time around, we’re going to change that. You might be thinking: I’m really interested in mapping but I don’t know much about companies or how to map them? A few months ago – I said the same. Now, we have a relatively straightforward way of mapping companies. Let’s look at the end product and then show how it will be done: Screen Shot 2015-04-27 at 10.12.25 In the spirit of full disclosure, let me tell you that this example was not put together by people. This data was scraped/extracted from SEC () and then visualised through our alpha tool: Octopus.

So how will we get to this pretty visualisation?

Step 1: Pick a company. We’ll help you with this step as we’ll have already chosen an influential company and bought all the corresponding fillings.

Step 2: Locate the ultimate controlling company through (1) Wikipedia (2) company’s website or (3) OpenCorporates. A useful way to locate this is to find the incorporation date and the legal entity name of the ultimate controlling company through Wikipedia and/or the Company’s official website.

Step 3: Find that company on opencorporates.com.

Step 4: Using the filings from the original company and any subsidiaries we already know about, scour PDFs for signs of other significant holdings and subsidiaries and repeat the process of research for them. Fill in the information on a spreadsheet which has been formatted specifically for this process.

Step 5: My colleagues will run a script and turn this into the pretty graph. Viola! We’ve got a straightforward road to mapping the networks but definitely not without challenges. It is paved with blocks, traps and mazes. Hope you’re excited about this adventure! If you are, here’s what we need you to do:

  1. Read about the data model of OpenCorporates.
  2. Have a read of the operations of G4S, Sodexo, Aviva and Capita..

If you’re a coder, it would  be great if you could help researchers extract data from the large PDF tables which often crop up. At OpenCorporates, we’ve used a few approaches but need your help to find out and document the ones that work best. If you just want to turn up and write a bot – don’t worry; we’re still steaming ahead with Map the Banks and need more bots! In fact, if you can finish a bot this weekend and then turn up to our #FlashHacks event – then you get a chocolatey gift.

If you’re not in London and want to attend a #FlashHacks event? You can join us online through our Slack channel. Email community@opencorporates.com to get an invite.