Who Runs Hong Kong?

Who Runs Hong Kong is officially live – an interactive visualization of 4,000 key people and 2,000 companies and organizations in Hong Kong and how they’re connected.   There’s more information here.

It’s as close to a structured journalism product as you’ll get – we just need to get the newsroom hooked up to updating the database on a daily basis, and then we’d be well on the way. That may take a little time, but the real value of WhoRunsHK is that we have a working demo and pilot project that allows us to work on enhancements as well as understand how users might approach a product like this.  It gives us real feedback so we can iterate.

Kudos to Irene Jay Liu, who ran this whole project, with much help especially from Chris Ip, as well as Malik Yusuf, Vivian Li, and interns Eldes Tran, Vanessa Ko, Nathan Griffiths, Stephanie Kwan from Hong Kong University’s Journalism and Media Studies Centre, and to developer Chris Spirito, who modified the ThinkMap program to get us a smart, custom visualization.  It’s not bad at all for a project pulled together pretty quickly over a well-designed database that gives us a a real range of stories to chase.

A bit more about the project:  It’s often said that Hong Kong is run by a relatively small elite – a group of people who have deep family, school, work and other personal connections.   What WhoRunsHK allows users to do is explore those connections and relationships.  All of the data in the site is from public records, so there aren’t any smoking guns or deep secrets in there; but simply visualizing it unearths all sorts of new value.

The idea harks back to my time at the Journal, where I pitched a similar project that I called People Maps; the goal being to take a smart visualization akin to what Muckety does but integrate the database maintenance into the day-to-day workings of a newsroom – the core of what structured journalism is all about.

In this case, we’ve built it off public records and the sweat of some very hard-working interns; but the theory remains the same: If we have a database of relationships of key people and companies, add some generally-known-but-not-easily-accessed (or not-so-generally-known) information, such as family ties or schools attended, and then have journalists update the database whenever they file stories on the people and companies, then after a while you have a monster database that’s increasing in value everyday – and can’t easily be replicated.

The real value, of course, is in how often and regularly the database is updated, and what “secret sauce” of not-easily-obtained information is built into it (otherwise someone else can simply replicate the site).  There are some elements of that here, but it’s still a work in progress.

When it does all come together, it will be a way to extract more value out of journalists’ work on a day-to-day basis; and it’s a way of giving longer-term value to the daily work that they do.  What’s not to like?

True, WhoRunsHK is free – where’s the money here?  There are certainly potential ways to go: One is to put it all behind a paywall, assuming people get used to it and value it.  Another is to put key parts of the database behind a premium subscription wall; really deep searches and visualizations will be restricted.  Another is to do customized searches for people who want much deeper research – while the visualization is good, it only surfaces a subset of the information in the database.  And yet another would be to allow customers to rent computing time on the database, letting them do their own custom research.

It’s not clear which or any of these models would work.  What really matters is seeing how people react to the information and track their interactions, and then iterate to figure out what might work.  It may be that this data isn’t tremendously valuable to a lot of people – although I suspect it is.  But certainly some data is – a relationship map of people in DC, or in China, or in key markets.

And much more importantly – if set up right, and with a newsroom structured to integrate into the product – the incremental costs of doing this are minimal.  So you get a product that has potential new revenue streams; you serve readers and the public interest; you don’t add hugely to newsroom workload; you gain a reporting tool; and you do it at low cost.

What’s not to like?