Comments on: Web Measurement

By: Becks

Becks — Wed, 04 Feb 2009 22:54:45 +0000

Level two? LEVEL TWO? bloody cheek

By: Paul Stallard

Paul Stallard — Sun, 25 Jan 2009 19:06:58 +0000

As you say the art of measuring emerging media is a difficult one, but I think your work here is very interesting and something that I will look at in greater detail over the coming weeks. Good to see this post listed in PR Week as well.

By: Brando Social» Blog Archive » Measuring influence is complex, not complicated

Tue, 20 Jan 2009 14:51:33 +0000

[...] Jed Hallam at Rock Star PR has been wrestling with how you might work out influence and struggling away with a web of influence. It’s a good effort. [...]

By: Brando Social» Blog Archive » Measuring influence is complex, not complicated

Tue, 20 Jan 2009 14:35:06 +0000

[...] been wrestling with how you might work out influence and struggling away with a web of influence. It’s a good effort.And we’re going to need something relatively simple (as opposed to complicated… stay with me on [...]

By: David Cushman

David Cushman — Tue, 20 Jan 2009 12:03:06 +0000

Interesting co-incidence with my attempts to think through how a google adsense model may calculate the value of your mentions in conversational marketing.
Will link back to this piece. dc

By: Richard

Richard — Mon, 19 Jan 2009 12:15:05 +0000

influence is a word that’s definitely banded around far too much in the industry at the moment, and as far as I can see, it’s only going to be detremental to evolving a pr and marketing arena that clinets trust, if start-up social media oriented companies keep shouting about ‘getting influencers on board’ and such. Maybe 2 years ago, but not now, you know, influence is entirely subjected.

I believe people can have ‘reach’ online, and a selection of metrics can be utilized to help show a client why they should be talking with a select group of people, because they have reach into their online world. But to suggest you can directuly calculate a persons ‘influence’ is madness, and if it keeps happening will blow up in all of our faces eventually.

@mat i like the ideas you’ve put up there

By: Mat Morrison

Mat Morrison — Sun, 18 Jan 2009 00:22:15 +0000

Like the idea. We’ve been working on network influence metrics at Porter Novelli for a while — the three that we’re beginning to get together on are:

1) Authority (“degree centrality” or “eigenvector centrality” — both on Wikipedia): a measurement of how many people (and who) link to a source.

2) Popularity: how many people read a given source (these numbers are almost impossible to get hold of without asking for them

3) Betweenness: does the source form a bridge between two groups that would otherwise not know what the other was up to?

For 1 and 3 we have a number of tools (including Technorati) — but the one that we work with most often is Rufus — which we’ve been developing for a while now.

The other metrics we look at are Recency, Frequency, and Tenure.

None of these metrics (yet) replaces a combination of desk research and interviews if we’re trying to assess sources in a new environment, and none of them replaces regular reading as a way of keeping abreast.

But they can help us justify our findings. Then we can plot them on the web =)

By: Ged Carroll

Ged Carroll — Sat, 17 Jan 2009 21:58:41 +0000

A surrogate measure may be akin to Page Rank: ratio of inbound links versus outbound links. Those that are more influential are likely to drive agendas and consequently be referenced. So this ‘page rank’ would relate to the radius of the sphere you describe.

But all this is conjecture, so keep the tongue firmly planted in your cheek.