A couple of weeks ago, I asked a bunch of people to send me their OPML files (for those of you who aren’t aware, an OPML file is what tells your RSS reader what feeds you’ve subscribed to — it can act as a way of moving your subscriptions between readers.) Some of the more trusting among them agreed, and that gave me the raw material for the first bit of my experiment.

Some red herrings

Along the way I uncovered a couple of things that were interesting but not (entirely) relevant to the experiment.

1. Some people are cagey about sharing their list of feeds: whether they consider it intellectual property, or whether they think that it may be too revealing, I don’t know.
2. Lots of people said things like “oh — my RSS reader? Haven’t looked at that in a while. I get all my news off Twitter these days.”

So what’s the experiment about?

You’ll probably know that I’m interested in networks of people and how information flows through those networks. I’m also interested in things like influence and whether and how we can identify and track its effects.

I’d heard about a word of mouth marketing campaign which was set up along the following lines: a sample set of pupils at a given school were asked “who’s the coolest kid in the school?” Clearly some names came up more often than others, whereupon the researchers went to those kids and asked them the same thing. By the end of the process, they had a good idea of who might be the key influencers.

This seemed like a good sort of thing to be doing. It was a simple idea, and apparently easy to execute. Everything else I was working on (the citation analysis and the network analysis in particular) seemed to be complementary.

A little history

So last year we did a version of this experiment for a client. We emailed and phoned a whole load of journalists and analysts whose beats covered our client’s interests and asked them:

Who do you read on a regular basis?

Who (other than you) should we be talking to?

If you were looking for information, where would you start?

Then we approached the people that they’d recommended and asked them the same questions. The purpose (as I’m sure you’ll have guessed) was to create a network map of who was whose go-to guys and girls. We’d take the data that we collected, push it through the usual processes and hey presto! we’d know who was really influential.

But what we actually found was that the journalists and analysts we asked seemed not to have any specialist sources. The people whom they cited were (in no particular order) their colleagues, the companies they covered (and their public relations representatives and agencies). Oh, and Google. Google was cited by everyone.

There were two obvious explanations for this:

1. It’s the truth: they really didn’t have any better sources. We’d all read about the collapse of journalistic standards; perhaps we were encountering it at first hand?
2. The privacy and intellectual property argument: maybe the journalists and analysts were constitutionally loathe to reveal their sources?

All my training and experience, however, points to the following reason:

3. Poor questionnaire design led to an inability to think (when put on the spot) of names and sources whom they believed were their influencers.

This is less unlikely than it may seem. If I were to ask you who your big influencers were, would you be able to answer?

All in all, it was a deeply unsatisfying exercise. We had envisioned what wild success would look like and this wasn’t even close.

Where we are today

I thought that it might be easier to ask people for their OPML files than it had been to ask them who were their influences. This is the equivalent of — say — asking to see a musician’s CD collection, instead of asking them about their musical influences. It’s not necessarily more accurate, but it might help expose a different picture.

Seven of my friends, colleagues and acquaintances sent me their OPML files, and that was enough to get started.

Early results

Between them, the first seven people subscribed to just over 1.5K RSS feeds, which gave me a lot of data to process. Here’s a picture of the network before I started processing the data.

You can see (I think) pretty clearly that there are several blogs (or RSS feeds — I’m using them interchangeably here) in the middle of the map that are linked to by several of the respondents. And there are some (around the edges) that are linked to by a couple of respondents. And there are lots that are linked to by only one respondent.

At this stage we’re interested in “indegree” — or “the number of OPML files in which we find RSS feed X.” Clearly the great majority (just over 1.3K) of the RSS feeds are single hits — that is, they appear in only one respondent’s OPML file. But that still leaves 180 RSS feeds with multiple hits:

Those 180 RSS feeds with multiple hits left us with a nice (and fairly predictable) distribution that looks like this:

Clearly the first-generation results aren’t very meaningful or accurate. Most often cited were author Seth Godin, ex-blogger-turned-lifestreamer Steve Rubel, and PR blogger Stephen Davies who were each read by five of the seven respondents. But more than half the respondents read my blog. Despite what I’d like to think, this is clearly an artefact thrown up by the sampling frame (my friends, colleagues and acquaintances).

Where next?

I’m hoping that we can extend this out a generation — and then keep iterating. I’m going to approach everyone with an indegree of 3 or above, and see how many of them will send me their OPML files.

I don’t hold out much hope of getting OPML files from people like Messrs Godin and Rubel, but if I don’t ask, I’ll never know, will I?

The current approach to WOM is to try to stimulate positive WOM while addressing or countering negative WOM. A sort of “accentuate the positive, eliminate the negative and don’t mess with Mr In-Between” strategy.

But what if we could do it a different way?

This idea stems from a conversation I had back in February with Martin Kelly and Andy Cocker of Infectious Media. Since that time I’ve chatted it through a couple of times with various interesting people. It’s not properly thought through yet, but following a chat a couple of weeks ago with Ketchum London’s new Head of Digital, the excellent Fernando Rizo, I’ve decided to put the idea out into the public domain to gauge what (if any) interest there is and whether I should continue to work on it.

“Word of Mouth” is hard to do well

I’ve read lots of word of mouth marketing case studies (there’s a great list over at WOMMA) and it strikes me that WOM is hard to do well for a few reasons. I don’t want to go into these in too much detail, but here are a couple of the structural issues:

1. Unless I’m a journalist, an A-list blogger or media personality or have some kind of platform, I probably have a very low reach.

   Despite everything pointing towards personal contact being the best impetus for positive word of mouth, most word of mouth campaigns compensate for my low reach by trying to get me to self-service my relationship with the brand and the campaign.

2. “Viral” distribution just doesn’t work the way most people seem to think it does; and this is particularly true when it comes to WOM.

   While I’m quite likely to tell stories about my personal experience of a brand and fairly likely to tell stories that involve a mutual friend, I’m much less likely to tell stories about other friends’ experience, and not likely at all to tell stories about friends-of-friends.

   Furthermore because of the ‘clumpiness’ of most people’s social graphs, geometric progression (the “I tell two people and they each tell two people and so on” effect) just doesn’t happen.

Homophily

One of the many reasons that WOM works is a thing called homophily — which roughly translates to “birds of a feather flock together”, or “you can tell a man by the company he keeps.”

I’ve written about examples of this before: for example, my analyses of twittering US Congresspersons and Westminster MPs which showed that one can predict with some reasonable degree of accuracy the political colouration of any given twitter account based on their mutual friends and follows (if you want to know more about the methodology, it’s worth reading Robert Hanneman’s chapter on cliques and subgroups.)

But there’s another side to the homophily coin; the social pressure to conform to the group’s norms.

Why I bought an iPhone

Let’s take a concrete example of this: it often seems to me that everyone I know has an iPhone. I made a conscious decision a few years ago not to buy an iPhone, but I’ve finally succumbed to the pressure.

While it’s true that several people have take time out of their busy schedules to tell me exactly how good the iPhone really is, I’m far more affected by what I perceive as the omnipresence of the iPhone. Everyone, it sometimes seems, has one except me.

Of course, on those occasions when I head down to visit my family in rural Hampshire, I’m reminded of the obvious truth: most people don’t have an iPhone. But that’s not the way the world appears to me.

How about turning it on its head?

What if we could identify people who are under social pressure to buy our products — who are being influenced by what their friends are doing and saying?

Let’s say for example that Apple and O2 (iPhone’s carrier partner in the UK) could work out that a significantly higher proportion than average of my Facebook friends access the service using the iPhone client, and could target me with special offers and rates to push me over the edge.

Instead of looking for WOM influencers, why don’t we look for areas of high potential — and target those people who are likely to be receiving lots of WOM stimuli?

Twixdagen does for Swedish politics what Tweetminster does for British. Hampus Brynolf (@hampusbrynolf) just sent me a link to this map he’s pulled together for their blog:

You’ll need to click through to his blog post to experience and interact with the map properly.

Hampus says that he used aiSee to generate an SVG file which could then be opened in Illustrator to “search and replace” on shapes, colors and lines (which explains the good-looking graph.)

Using nothing more than their public twitter relationships, is it possible to predict whether a US Congressperson is a Republican or a Democrat? The answer seems to be a guarded “yes” — our tools predict correctly 40/46 times (or around 87% of the cases.)

This post follows on from a post earlier today in which I asked, “can we calculate party affiliation?” The data set in the earlier post was gathered from the 16 members of the UK parliament who are on Twitter and the relationships between them.

Tweetcongress maintains a list of US congresspeople on Twitter. Today (February 13, 2009) there are 76 congresspeople on the service, but when I collected my data set of “who follows who” on February 3, 2009 there were only 65. Of these 65, fully 19 (29%) lived a life of noble isolation with regards the network — none of their peers linked to them, and they in turn linked to none of their peers. Removing these Miss Havishams from the data set leaves me with 46 twittering congresspeople who form a network.

Now as both social network analysis and Aesop would have it, “a man is known by the company he keeps.” What I mean by this is that given the partisan nature of politics, we should expect that Democrats will link to other Democrat twitterers more often than they link to Republican twitterers and vice versa. So that’s what NetDraw[1] , the software I’m using for most of this stuff, looks for, or more accurately:

To identify factions, NetDraw software iteratively searches for a distribution of nodes among a selected number of factions to minimise the number of connections between factions and to maximize the number of connections within factions.

Whatever. So I let NetDraw loose on the data, and here’s what it did.

I coloured the nodes red for Republican and blue for Democrats[2], labeled the nodes by party (for the sake of clarity, and for the hard-of-thinking, that’s “R” for Republican and “D” for Democrat) then counted all the nodes where label said one thing but colour another. There were six of these nodes; so NetDraw got the answer right 40⁄46 of the time (just about 87%.) This is less than the astonishing 93.75% accuracy we got with the Westminster twittering members of parliament in the previous post. Nevertheless I think we can safely say that it’s not a particularly integrated (or bipartisan) network if we can predict party affiliation with quite such success.

Here’s exactly the same map with the errant sheep re-labeled with their proper names so it’ll be easier to refer to them (if it helps, you can click on the image to view or download a larger version.)

You’ll see, I hope, that NetDraw has made a pretty good fist of the job. Where it has gone wrong on the whole is where the data clearly suggests something else. So Rep. Jared Polis for instance follows (and is followed by) no Democrat peers. Rep. Nancy Pelosi (D) and Sen. Richard Durbin (D) follow each other, but since Pelosi is followed by several Republicans and none of her other Democrat peers you can see why the algorithm has made the incorrect guess that the two of them are Republicans. Long-serving member Neil Abercrombie, as discussed in a previous post on US Congress Twitter folk, forms a bit of a bridge between the two parties, so despite his membership of the Congressional Progressive Caucus and liberal voting record, from the Twitter network point of view, his affiliation is somewhat ambiguous.

Sen. McCain follows none of his peers, and appears to inherit his incorrect attribution from Sen. Susan Collins. For the life of me, I can’t work out what makes it think that Sen. Susan Collins is a Democrat. She really isn’t, you know.

Note 1: NetDraw is a free program written by Steve Borgatti from the University of Kentucky. If you’re interested in playing around with this stuff, you’ll need to get yourself a copy.

Note 2: Actually, that’s not true. Despite a friend sharing the simple mnemonic that “‘Republicans’ and ‘red’ begin with the same letter,” I just can’t get it out of my English head that the Republicans should be blue and the Democrats red. As a result I waste precious minutes re-colouring these maps in Illustrator. It is worth pointing out that I also have problems with “left” and “right” on occasion — preferring instead the binary opposition “left” and “No! no! The other left, for God’s sake!”

This is a follow-up post to Why doesn’t the Tory MP have Twitter friends? — a report on some early research into the interrelationships between the few Westminster MPs who are on Twitter.

According to Tweetminster, the number of UK MPs on Twitter has doubled since this time last month. Where there were eight Twittering MPs, there are now sixteen. Here’s the map that shows who follows whom (the labels may be too small to read — if you want to see a larger image, click on the map.

I’ve coloured each node to show party affiliation; for those of you who are unfamiliar with British politics, Labour (our left-of-centre party) shows up in red, Conservatives (our right-of-centre party) in blue, and Liberal Democrats (what it says on the tin) in yellow.

The size of each node represents the individual’s “betweenness centrality” — a network analysis term that helps us place a value on individuals within a network. To give you a sense of what it means, the higher the betweenness centrality of an individual, the greater the impact when you take them out of the network. For those of you who work in large companies, it may be worth noting that senior management’s personal assistants generally have very high betweenness — something that is mostly remarked upon when they go on holiday (simultaneous translation: “take a vacation”.)

So far so good. By now, regular readers will probably be kissing their teeth and thinking “so what?” I’ve done a lot of these Twitter maps in the past and the novelty must be wearing off on you by now.

So here’s the thing. There are a few network analysis techniques that let one identify cliques and factions. What we’ve got here is a small set where we already know what people’s affiliations should be. How interesting, I thought, would it be to see how well the calculated result fits the real world data? Here’s what I found:

To my fairly untrained eye, the fit seems remarkably good. Only one MP out of sixteen is incorrectly placed (Derek Wyatt should be tightly affiliated to the “Labour” faction, but shows up affiliated to the “Lib Dem” faction.) That’s 1⁄16 wrong (or 93.75% accurate.)

The usual caveats

The method I used relies on you telling it how many factions there are. I knew there were three, so that’s what I told it. But what if I’d said two, or four? The software I’m using tells you how well-fitted the result is to the data, so there’s some feedback at least. But I’m going to need to play with this a lot more before I’m fully confident about putting it in front of a client.

I’m not a statistician, but even I know there’s something fishy about my claims to 93.75% accuracy.

Colleagues and regular readers will know that we’ve been working on an “online influencer mapping” tool called Rufus. Those of you who’ve had a chance to use Rufus will know that it requires a seed list of URLs to get started. Creating this seed list can be automated in one or two ways, but one of the fastest, most effective, and most sensible ways to build a seed list is still to do it by hand.

We’ve got one or two other processes that also require us to build a seed list. No doubt other people do too — lots of web research is quite data hungry. So — because I’ve found myself telling a few of my Porter Novelli colleagues how we go about the process, I thought I’d share it here, in the interests of:

1. having somewhere to point people in future,
2. general good-heartedness: I’ve learned a lot from people in the past, and I like to give stuff back, and
3. getting feedback and tips from people about how they might go about the same process.

Oh – and while these methods should work in any language, please bear in mind that I tend to think and work in English. I’d appreciate feedback on how best to localize these methods.

Building a seed list: 5 easy methods

With all these methods, there’s no substitute for checking out the blog. I don’t ask people to read the blog (that comes at a different stage of the process altogether) but you should at least click through and see what you’re dealing with. In fact, method 3 rather relies on you visiting the blogs you’re researching.

1. Look for someone who has already done your research for you

Start by being optimistic. Generally you’ll find that someone else has created a list of the “top ten” (or however many) blogs in the niche that interests you. Take a look at Brendan’s regularly updated PR Friendly Index for example. If you’re searching for English language blogs then you could do worse than start by looking at Guy Kawasaki’s Alltop. But simply Googling for lists of blogs or blog charts should get you a long way.

This is generally a source of fairly high-quality data. One thing to watch out for, though, are search engine spamming link farms, and shady “Make Money Online” (MMO) directories. You’ll learn to recognize these soon enough, but as long as you’re visiting all the blogs you’re putting on your seed list you should be alright.

2. Do a tag search on delicious

I picked up this technique from Anthony Mayfield, who showed me that by searching on the delicious social bookmarking site for the tags “xxx” and “cool” and/or “inspiration” you could find sites about “xxx” that people thought were cool. Knowing what your digital trendsetters think is cool is one hell of an insight.

For our purposes though, we’re looking for cool blogs. So (1) click the “Explore Tags” tab on the home page, and then (2) type your keyword and the word “blog” into the search box. Couldn’t be simpler?

Well — actually it could be simpler. You can query the delicious database when you type the URL into the address bar of your browser like this:

http://delicious.com/tag/blog+keyword

Where “keyword” is the word you’re looking for.

When you get the results, check the ones that (a) have the right kind of title (if you’re looking for French blogs, look for French titles for example), (b) have the right kind of tags and description and (c) have been bookmarked most often

If there’s a better local language social bookmarking site, I’d use that whenever possible. For example, Mister Wong is a good one for German language sites.

A quick note: social news sites like Digg and Reddit, and “serendipity browsers” like StumbleUpon tend not to work so well in my experience.

This method also owes a lot to Marshall Kirkpatrick. You might like to try out the Yahoo! Pipe that I built based on the process that Marshall documents.

3. Look for blog rolls

On every blog you visit during the research process, look for the blog roll — and check the likely-looking links. See if they’re useful or useless. Quite often you’ll find that someone who has an interest in widgets will also read and link to blogs that cover widgets. That, after all, is the principle on which Rufus works wrote small. So we reckon it’s a pretty good approach.

4. Ask your Twitter followers

Seriously — this works. Well — it worked for me and my team from around +100 followers onwards. I’d be interested in others’ experience.

5. Call someone

Get hold of someone who knows about the subject and phone them up or get them on IM. Category experts are an excellent source of low-volume but high-quality information. It’s time consuming, but can work well if you have the right contacts. Journalist friends might be a great source of blog lists.

I’ve purposefully left this one till last; I think it’s a good rule of thumb to do your desk research before picking up the phone. That way you can ask intelligent questions instead of damn fool ones.

Using a text editor

I try to keep two lists running all the time that I’m working; a scratchpad list of blogs I have yet to visit and the seed list itself. Because I’m on a Mac, I use the excellent BBEdit (there’s a free version called TextWrangler which will be just as good for most people.) If — as is more probable — you’re on a Windows machine, you might like to try the very powerful but slightly less pretty Notepad++. But if you just want to use Excel, though, that’s fine, too.

Three weeks ago (and at the prompting of my colleague Eddie Garrett who heads up Porter Novelli DC’s digital team) I mapped out the interconnections between US Congress Tweeters. We’d been working on a Twitter crawler and it seemed like a good opportunity to test things out on a new data set.

This is a follow-up post. Once again it was prompted by a third party: Christie Findlay at Politics Magazine asked whether it would be OK to print a copy of one of the maps in their March edition. I’ve heard that three weeks are a long time in politics, so I thought I’d better run the crawl again just in case. Also I’ve got a new crawler that uses the proper Twitter API (I can see some of your eyes glazing over you know. Just skip ahead when that happens.) I’d tried it out on the Porter Novelli data set, but welcomed a chance to try it on something more meaty.

So yesterday morning before work I ran the crawl. I use the excellent Tweet Congress as my source of information about which congress people are on Twitter.

Here’s the data I get from them republished as a Google Spreadsheet.

I use Twitter itself as the source of information about who follows whom and who — in turn — is followed by whom. Then my computer does a little analysis on the data I get so that I can put it through our mapping software. Here’s the map. It’s a little too small to see, but if you click on it, it’ll take you to a huge version on Flickr.

It’s probably worth going through some of the features again briefly, just in case you haven’t seen my other maps. Actually, I don’t think I’ve ever really explained them properly.

The arrows show the direction of the relationship. If Senator Susan Collins follows Senator John McCain then there’s an arrow that points from Sen. Collins to Sen. McCain. If Representative Jason Chaffetz (Rep. Utah) and Representative Jared Polis (Dem. Colorado) follow each other then there’s a double-headed arrow between Reps. Chaffetz and Polis.

If they follow none of their peers and none of their peers follows them, then they’re relegated to the bench. In this map the bench is the list of unconnected circles in the upper left corner. For those of you who are comparing this to the last map, one of the interesting things to note is that Senator Hillary Clinton has left the ranks of the Democrat congressperson Twitter folk.

I’m not sure why Rep. Thomas Price (R, GA) and Rep. Steven Driehaus (D, OH) have Twitter handles listed. They aren’t on Twitter as far as anyone can tell. Perhaps it’s a tyop?

So far, so obvious I suspect. And it should be obvious to any US readers, the red circles are the Republicans and the blue circles are the Democrats. This is mildly contrary and confusing to my UK audience who are used to the left-of-centre Democrat analogue Labour Party being red and the right of centre Republican analogue Tory party being blue. It’s like going to France and forgetting that the tap with C on it gives you hot water. For my US readers, you probably know a tap better as a faucet. It’s rather exhausting having to do all this translation.

You’ll notice that the circles are different sizes. If you’re very observant, you’ll probably notice that the large circles have more arrows pointing at them, and that the smallest circles have no arrows pointing at them. For this map I’ve sized the circles by a factor called “indegree” which pretty much translates as “the number of arrows pointing at you.” Indegree is used in network analysis as an indicator of popularity or authority (depending on the circumstances.)

First caveat

This can be pretty spurious when you’re looking at what is (after all) only one communication channel for this network. We have to imagine that congresswomen and men (although it’s mostly men at present) will have opportunities tp communicate face-to-face, on the telephone, and via email, memos and mail and the like. So I have no idea how well this fits the reality. I’d have thought (for example) that Sen. McCain was pretty authoritative — but the Twitter map shows him to be peripheral.

This leads us to our second caveat.

Second caveat

For most of this group, the primary audience is not their peers. It is (rather obviously) their voting public. As we’ll come to see some of them have established some pretty hefty audiences (although none of them is really in celebrity territory yet.) So that probably makes this network even less representative.

Still it’s an interesting exercise. Let’s carry on.

I don’t want to labour this too much. Those of you who regularly read this blog must be getting sick of hearing about betweenness centrality. For those of you who don’t know, then here’s a brief explanation: betweenness is a measure of how important the person is in terms of flow of information. It’s calculated by working out who is on most shortest paths through the network.

If Twitter were the only way that information flowed in Congress, then Representative John Culberson is (still) the single most important person there. Taking him out of the loop would mean that the flow of information would slow down considerably, and that some people would never get the message at all.

Take both him and Representative Neil Abercrombie out of the picture and it’s like something out of West Side Story.

Sharks and Jets

Because every social media expert other than me has written a term paper on “How Obama used social media to win the election”, and because the general social media buzz tends to be slightly Democrat biased we’ve become accustomed to the idea that the Democrats do social media well. That may be true. But all I can say is — looking at this map — you’d never know. Republicans outnumber them 1.6 to 1 on Twitter (once we discount the spectral accounts of Driehaus and Price.)

I’d like to say that the Democrats had made up some ground recently, but they just haven’t. A little more analysis shows the growth in Twitter adoption by two parties. Democrats made a good start, but the Republicans really grew their numbers during June and July last year.

update: I’ve fixed yesterday’s chart. I got the whole Red vs Blue thing the wrong way around again. Thx to EG for setting me straight.

However the new Democrat joiners aren’t really linking to each other. Why is this? If you’re hoping to attract followers, one good strategy is to make the most of your network: their followers are more likely to be interested in what you have to say. More interested, that is than an equivalent control group chosen at random from the Twitterverse. So not joining networks just makes things harder.

Partly of course, we can attribute it to naivety. Most new people coming into blogging, Twitter and the rest see creating content as the hard work. Most people who’ve been doing this for a while know that it’s the promotion and moderation that sucks up your time.

Perhaps we might also attribute this to the point raised in Caveat 2 above — that they’re mostly interested in their electoral constituencies. However, if this were the case then I think we’d expect to find factions (or “cliques” or “clusters”) between congresspeople of the same colour who represent the same state. Nothing obvious there, though (and the data set’s quite small). So let’s stick with the “Twitter newbies” theory for the moment.

How many people follow the Congressmen?

So far we’ve only looked at peer relationships: all the charts and calculations we’ve done so far are based on a much-reduced data set. Experience shows that it’s very difficult to map “all followers” (we end up with something that looks like a hedgehog that’s had a wild party in a cranberry field.) But we can still count followers. Here’s a quick list — you can see that there are one or two superstar congresspeople, but that the rest really aren’t doing even as well as your average journalist.

OK — now let’s assume that these aren’t unduplicated audiences; that there must be several people interested in US politics who follow more than one, many, or all the names on the list. What is that duplication like?

Luckily I’ve just written a little analysis script to look at exactly this sort of question, so I can tell you with confidence that (while the absolute numbers above might lead you to think that the congress Twitterers reach an audience of around 56K) if they all pull together, they actually reach an unduplicated audience of 24K. Of course, now I look at it, I realize that they are unlikely to pull together in this way, and that I should have separated the lists into Republicans and Democrats before I went at this. This is somewhat frustrating — but I can cover it in a new post, I suppose.

Here, then, is the somewhat useless Pareto chart showing the unduplicated reach as though the US were a one-party state.

I’ll come back to this, but if you have questions, you might like to look at this Q&A I wrote over the weekend.

Three days after my last map, and after lots of internal nudging from our CMO Marian Salzman, her two helpers Tikva Morowati and Zeenat Duberia and local activists like Juriaan Vergouw, Burçu Kaptan, and Umut Ersoy, the map of Porter Novelli people on Twitter looks very different. (You can click on any of the maps in this post to go to their Flickr page where you can choose to see them at larger sizes.)

This is a map of the 156 Twitter people that I knew about this morning (there are still others coming in). The size of the blob is an indicator of how many of that person’s peers follow them on Twitter, and the colour (running from white at the “hottest” through orange and red to blue at the “coldest”) indicates how structurally important that person is. For more on this, see yesterday’s post showing the Twitter map on 17th Jan 2008.

More cohesion

Here’s what happened when I remove the ten main connectors from the old map. It falls apart.

17th Jan map (67 nodes)	10 highest betweenness nodes removed

And here’s what happens now.

20th Jan map (155 nodes)	10 highest betweenness nodes removed

You’ll see that much less damage is done now; the network is much more decentralized, thanks to a concerted effort by Marian et al. to promote cross-linking behaviours. Of course, we’re now removing around 6% of the network compared to 14%, but if the network were still as unlinked as it used to be, that would make no difference. This is on the whole, a more cohesive network.

I say “on the whole” because of the isolates (unlinked nodes) that are displayed at the top left. These are people who are new to twitter, and who haven’t yet made any friends or attracted any followers. This is one of the problems we’re trying to address; I’m considering building a “follower bot” that will help new Porter Novelli joiners (and anyone else who cares to do so) automatically follow all the other Porter Novelli twitterers (should I get over my distaste for the word “tweeple”?)

Weaknesses

As Tikva and others have pointed out, simple linking isn’t enough to really rank people in this network. You’d want to look at other quantitative metrics. Recency, frequency and tenure spring to mind, of course; but I’d also want to see things like who (in the network) is most often addressed or quoted using the “@<username>” convention. We’ve been building an “eavesdropping” bot to help us get a grasp of that sort of thing, but we’re still not there (the code works, but I can’t get it to run on my server yet…)

On the qualitative side, we’re also missing important things like “how much one person likes, is indifferent, or dislikes another” (there has to be a better way of expressing that.) I don’t really do qual measurements at present; I want to be able to get a whole load of stuff out of the way before I even begin to deal with tricky stuff like that. Is this an Aspergers thing, do you think, or am I just being male?

Marian Salzman (our Global CMO here at Porter Novelli) has had the inspired idea of getting people in the agency to tweet about the most exciting story this week (probably) — the inauguration of Barack Obama

You can see the results of the experiment on her blog.

I’m all for this, of course, for several reasons:

1. It gets new people onto Twitter
2. It helps us create a stronger network among Porter Novelli twitterers
3. It means I can track who at the agency is on Twitter

Simply asking people doesn’t really work. Marian’s scheme gives us a shared sense of purpose, which is one of the criteria for creating a strong community. It’s a much better plan all round.

But — of course — I want to see whether I can track how that works in practice. What do the before and after photos of a strong community look like?

The map above is the “before” map. You can see that there’s some nice heavy clumping down in the bottom of the chart, but a lot of pendants towards the top of the graph — people who are held into the main network mainly because of @robmctree. If he were to leave the network, you can see that some of them would float away to become like the eight other unconnected nodes at the top left.

That’s why his node shows slightly “hotter” (we rank the nodes according to a colour scale from deep red at the coolest to white at the hottest.) The hotter the node, the higher its “betweenness centrality” (basically a measure of its structural significance – if you remove the nodes with high betweenness centrality, the network tends to fly apart faster than if you remove the others.)

@mediaczar (that’s me) is big and white. The size indicates how many other people in the network follow that user (me.) Normally the big nodes in a network have all the betweenness centrality.

What’s interesting is using this technique to identify those (like @robmctree, @noahbanning, @volapuk, @angie_s and @amytokes) who have relatively low link popularity but high structural significance.

To give an idea of how significant these high betweenness centrality individuals are to the network, here’s are two maps: the original map, and a map with the ten highest scoring individuals removed.

Before…

and after.

Like finding “top bloggers”, finding “top twitterers” is relatively easy. But approaching them is harder: there may be more demands on their time, for one thing. But if we can identify those who are less “popular” but who may still carry lots of weight within the network, we may be able to find softer targets for our comms activities.

On another note: this project would never have happened without @mariansalzman, @tikkers, or @zeenat58 who have been tireless on the internal email channels. This only goes to show that mistaking the channel for the network is an error into which we can too easily fall.

Public service message: please don't push strangers in front of oncoming trains (by freshelectrons on Flickr)

When we’re using push channels like display ads, direct marketing or pull channels like websites or search marketing — numbers are what count, and numbers are enough. But when we are talking about social media channels, we shouldn’t target strangers. Instead, we should look at our existing relationships and learn how to make the most of these to our mutual benefit.

I don’t know whether you’ve had the experience of meeting someone famous in an ordinary context (in the street, say, or in a supermarket queue). I have.

It is a profoundly disturbing experience. For a split second your brain tells you that this is someone familiar but not why. Since you’re not expecting to meet David Bowie in your video store, your brain leaps to the most probable conclusion — this is clearly an old acquaintance or a friend-of-a-friend. By the time you realize who it is, you’ve already been staring at them too long, possibly waving and beginning to say hello.

What’s unnerving about this experience of course, is the asymmetry of the relationship; you know who they are (and possibly even some intimate details of their private lives) but they have no idea who you are. For all that you think you know them, you are in fact complete bloody strangers.

The circle of complete bloody strangers

At Porter Novelli, we’ve been trying out a new way of helping people think about the targets for our social media activities. Targeting in social media is one of the many places where conventional marketing experience fails to help; and indeed, generally hinders. For want of a better name, I’m calling it the “circle of complete bloody strangers.”

This may not be the most elegant name, but it does help simplify the problem. Here’s the synopsis: in traditional on and off-line marketing disciplines, planners like me tend to play a numbers game. We’re looking to reach the greatest number of people who will be receptive to our message, and (whether behavioural or demographic) we define our targets accordingly. And so, we spend nearly all our time crafting our messages and campaigns for complete bloody strangers.

Targeting strangers

Agencies do everything we can to get to know our target audience. At Porter Novelli, for example, we follow a process that has us (among other things) conducting desk research, running focus groups and interviews with representatives from the audience, and doing mystery shopping. At one of the wackier ad agencies where I used to work, there were workshops where hypnosis was used to break down barriers and aid rôle-playing exercises.

But however well we get under the skin of our target (say “C2DE mothers with young families”), they remain complete bloody strangers.

You might recognize them in the street, say, or in a supermarket queue – but like the celebrities whose lives you follow in the news – they have no idea who you are.

Social media marketing and social currency

There is (to my knowledge) no single precise definition of social media marketing. Many, I think would agree that it involves something like “using blogs and podcasts and social networks and media sharing sites and stuff to promote a product or service,” but that’s hardly precise.

Some would argue that simply buying advertising on a blog/podcast/social network meets the requirements, others would say that it doesn’t.

As a broad digital comms planner, I tend not to fall one way or the other, but suspect that if buying media or sponsorship was all that it took, we should all have stopped worrying and gone home by now.

I suspect that when we talk about social media marketing, we mean something like “generating word-of-mouth buzz and recommendations – whether for a product or a campaign promoting that product.”

I believe that we’re trying to use existing social relationships between colleagues, friends, tribes, and niche interest groups as the pathways along which we disseminate information. This isn’t as sinister as it sounds; we know that people use brands and branded content as what Douglas Rushkoff calls “social currency.” Content, he says

…only matters in an interactive space or even the real world … because it gives us an excuse to interact with one another. When I was a kid, we’d buy records not solely because we wanted to hear whoever was on them; we wanted an excuse for someone else to come over! “What are you doing after school? I got the new Stones’ album…”

In this sense, our content choices are just means to an end – social currency through which we can make connections with others.

So — in Rushkoff’s terms — what we’re doing when we do “social media marketing” is stamping new coins that can be released into the general circulation.

At Porter Novelli, my mantra is that “people want to promote themselves, and may be prepared to let you help.” This helps me remember that, however important my client’s campaign appears to me, my audience’s motivation is markedly different.

Why we shouldn’t talk to strangers

Take a look at the target diagram again. Normally I scrawl this on a flip chart and tell the story as I’m going along. The circles aren’t precise, and I often draw them differently depending on context, but these things are consistent: you are always at the centre of the target, and the outer circle is always defined as “complete strangers”.

Now ask yourself – who’s most interested in what you have to say, and who is most likely to repeat it in subsequent conversations? I’d say that the progression goes something like:

1. You (naturally.)
2. People you’re close to (friends, family, and colleagues for example.) If you can’t get these people interested in what you have to say, what hope do you have?
3. Your “ecosystem”; people who depend on you economically – whether suppliers or employees or retailers who carry your product or developers who develop for your platform. Trade and specialist bloggers and journalists fit in somewhere around here.
4. People who have a strong relationship with your product, tailing off from advocates to simple repertoire purchasers (people who regularly — but not exclusively — buy your product.)
5. Complete bloody strangers.

Let’s take an example.

1. The people who will be most interested in the launch of your new FMCG ad campaign will be people who work with you and at the agency who created it.
2. People at other agencies will also be interested in the campaign, so it’s a good thing that there are lots of inter-agency relationships. And so will your friends and family — they like seeing what you’ve been up to and will tell their friends.
3. The retailers who stock your product will naturally be excited to hear that there’s a campaign coming. And Campaign and Ad Age are always hungry for content.
4. If you’re smart, you’ll have run test campaigns among your customer database of people who have asked to receive information. Asking their opinion makes them feel involved, and means that they’ll feel a little pride-of-ownership when the campaign does launch. “I helped with that,” they’ll say to themselves, and maybe to their friends.
5. If your media plan and creative are any good, you’ll create a stir among the complete strangers.

Do you see what I mean? As you move further and further out, the more tenuous the connection, and the less certain the response.

Another thing to bear in mind; the further away you are from the middle of the target, the more noise you encounter. By the time you’re out in the complete strangers circle, you’re just another voice making demands on their time.

So why do so many of the campaign ideas I come across require the active involvement of complete bloody strangers for them to be successful?

Let’s host a competition on YouTube where people send in films about our product! Wheeee!

Let’s create a community! Wheeee!

Let’s send out a short film and hope that people will send the link to each other! Wheeee!

Let’s make a Facebook app! Wheeee!

None of these ideas is inherently wrong from the traditional marketing point of view. As an industry, we are (rightly perhaps) judged by total exposure to our campaigns. For most channels, numbers are what count, and numbers are enough.

But how on earth can we hope to influence the social networks of complete bloody strangers when we aren’t even making proper use of our own?