Building Design presents subjectively sorted Twitter list

Some 48 hours after The Architects’ Journal published its #AJtCn100 of ARB-registered architects (post), rival weekly Building Design has delivered its own top 100 influencers in the world of architecture.

The BD Twitter 100, however, is not published as a league table or ranking (though there are a couple of top tens: of UK architects and of UK architectural practices). Instead, because the BD team reckon PeerIndex is not yet sophisticated enough to give a truly representative view of architectural influence, it is openly presented as a subjectively ordered list. I am pleased to see Su Butcher at the top, just above Mel Starrs and Hugh Pearman, editor of UBM sister title RIBA Journal. @EEPaul ‘lurks’ down near the bottom, but then I am not an architect and architecture is only one of my many interests, and BD has adjusted its listing accordingly, as it explains:

… PeerIndex is still in its beta phase and the final ranking is far from perfect. … users who only occasionally talk about architecture but talk a lot about other subjects can rank higher than those who are more useful professionally.

The diversity of the topics PeerIndex looks for is also unhelpful when looking for authority around a specific subject – an individual’s top-five benchmark topics can range from baseball to design. Some users have reported that their benchmark topics bear little relation to the subject of most of their tweets.

It isn’t difficult for someone to boost their ranking by repeatedly tweeting or posting around certain subjects without contributing useful content.

Until PeerIndex resolves these issues, it remains potentially confusing, which is why BD has also used subjective human judgment to create its list.*

These perceived shortcomings in PeerIndex will diminish over time as the software engine is developed further. Indeed, earlier this week, I watched a tCn-produced video of Part One of an interview conducted by Ryan Briggs with PeerIndex founder Azeem Azhar:

Towards the end of this video, Azeem responds to a question I posed: Can groups be ordered by specific topics? He says:

“We don’t have the ability to look at a group ordered by PeerIndex score within a topic right now, but it’s coming very soon and certainly by the end of summer you will be able to do that.”

This development (which I first asked PeerIndex about in March) will not only enable more accurate assessment of influence within a profession, it will also enable people to find others who share their key interests in particular topics. Architects might, for example, be able to target their networking on fellow designers who are focused on niche areas such as health or education, or who use particular types of software, or who favour timber frame construction, etc, etc. If such an advance can also be matched by tweaks regarding re-tweeting of content and recency of tweeting, then we will be a few steps closer to being able to use Twitter to help gauge opinion leadership.

* Update (2030hrs, 1 July 2011) – While the #BDTwitter100 has been manually sifted and rearranged, the crude PeerIndex listings, both the ‘Architecture 100’ and the longer list from which it was derived (currently 464-strong), can still be viewed online.


  1. Good post Paul. Subject lists would take a lot of the leg work out of the lists and enable more targeted lists – construction is a big old industry.

    1. Thanks, Alex. That’s a view I’ve held since I first looked at PeerIndex in detail. When the first tCn/Construction News list came out, it included an immense range of people from right across the built environment spectrum. An individual may be ‘influential’ on, say, BREEAM or PassivHaus, but their apparent “influence” might be largely irrelevant to someone working as, say, a demolition contractor or a construction lawyer. We need PeerIndex or Klout or someone to give us ways to narrow the sphere of influence so that we can make meaningful assessments.

      I would also welcome the development of some industry- or profession-specific “Topic Fingerprints”, to see who might be broadly influential across versus those who might purely focus on particular quadrants. Looking at the existing Fingerprint, it should be relatively easy to develop a construction-oriented version – MED might be about Health & Safety, for example, while SCI might be about sustainability, etc. Non-business dimensions (Sport; Leisure and Lifestyle) might be replaced by new ones – perhaps one on Legal/contractual, another on FM and whole life, for example. It might potentially also be linked to industry KPIs so that there could be a connection between individual interests and industry drivers.

Leave a Reply

Your e-mail address will not be published.