Vulture Culture | An Impromptu Link Sociology

Item: SEC Launches Inquiry Aimed at Private Equity.

Federal regulators have launched a wide-ranging inquiry into the private-equity industry that examines how firms value their investments, among other matters.

The Securities and Exchange Commission’s enforcement division sent letters to private-equity firms of various sizes in early December as part of an “informal inquiry,” according to the letter and people familiar with the matter.

In response, as PR Watch reports, private equity firms from all over the world have mounted a PR campaign using viral strategies to polish the tarnished public image of the sector as a flock of “vulture investors.”

My bookshelf contains the second edition of The Vulture Investors: The Winners and Losers of the Great American Bankruptcy Feeding Frenzy — published in 2000. There is also a really badly written financial thriller with a similar title.

At any rate, this mobilization of a PR rapid response team  presents a unique opportunity to track the development of a viral campaign from brainstorm to viral firestorm, in case of success.

A first step might be a simple crawl, with Navicrawler, of the principal pages of the campaign — its Web site, its Twitter page and its Facebook profile, as shown.

During the crawl, is possible to sort out and expunge the usual noise in the circuit — the Facebook page for every language there is, along with links to various other utilities and promotions, and the similar cluster repeated on every Twitter page. Or, if you prefer, this extraneous content can be edited out afterwards with greps and regexes.

The next step is saving an edges list in CSV format, then opening and editing it in a spreadsheet editor. Simplify the data as a spreadsheet of two columns — «Source and Target» — then save it as a Microsoft Excel .XLS file.

Import the resulting file into the yEd graph editor.

Visualize and edit the resulting network, making use of various functions yEd makes available.

The command

Tools > Grouping > Autogrouping > Natural Clusters

can be a useful step in starting to separate the wheat from the chaff and understand the underlying structure, if any.

This cluster is defined in terms of the internal tree structure of Spain’s daily El Pais, for example.

The autogrouping fares less well with the crawl data we developed for PEGCC.org, but does collect a lot of information about the firms and industry groups behind or tangentially touched by the campaign.

Compared with the analytic functions available in Pajek, yEd is only middling powerful and useful, so I often find myself saving a Pajek network, opening and editing it in Gephi, then saving it in a format readable by yEd. yEd, I find, offers more striking visuals.

As it turns out, autogrouping followed by brute force analysis of the common thread in the autogrouped elements makes for a nice list of the international activities of Innovations Media Consulting, probably the most successful — and under the radar — consultancies out there.

The link sociology — or ecology — of the P campaign makes more sense when sorted first with

Tools > Grouping > Autogrouping > Sub-Trees

and then hand-sorted into buckets including academic institutions, third-sector foundations and NGOs, industry firms, professional associations, media outlets, government bodies, and so on — it is fairly easy to filter out echo chamber components such as generic Facebook and Twitter internal links.