Attention + Influence do not equal Authority

In the dustup over whether it is a good idea to sort Twitter posts by authority – defined as the number of followers one has – John Naughton rises above the cloud to see a larger fallacy in the discussion: The number of followers one has does not equal authority. It stands for influence (or I’d say, it is a proxy for attention – and then, in some cases, influence).

The problem Naughton sees is the same one that plagues analysis of online discussion using media metrics. In mass media, of course, big was better because you had to be big to own the press: Mass mattered. We still measure and value things online according to that scale, even though it is mostly outmoded. Indeed, we now complain about things getting too big – when, as Clay Shirky says, what we’re really complaining about is filter failure. That is why Loic Le Meur suggested filtering Twitterers by their followers; he’s seeking a filter.

The press was the filter. And the press came to believe its own PR and it conflated size with authority: We are big, therefore we have authority; our authority comes from our bigness.

But the press, of all parties, should have seen that this didn’t give them authority, for the press was supposed to be in the business of going out to find the real authorities and reporting back to what they said. This is why I always cringe when reporters call themselves experts. No, reporters are expert only at finding experts. Now to put this back in Twitter terms: Reporters don’t have authority. They have attention and possibly influence because they have so many followers. But that doesn’t give them authority. There’s the fallacy Naughton pinpoints.

“So we need to unpack the concept of ‘authority,’” Naughton argues.

One way of doing that is to go back to Steven Lukes’s wonderful book in which he argues that power can take three forms: 1. the ability to force you to do what you don’t want to do; 2. the ability to stop you doing something that you want to do; and 3. the ability to shape the way you think.

In my experience, the last interpretation comes closest to describing the authority of the blogosphere’s long tail. It’s got nothing to do with the number of readers a particular blog has, but everything to do with the intellectual firepower of the blog’s author.

Naughton argues that the number to manage on Twitter is the Twitter_index – that is, the proportion of followers to (what?) followees. He believes it ought to be 1.0 – that is, equal – “otherwise one gets into the online celebrity, power-law nonsense that Le Meur describes.”

I wouldn’t go quite that far, but I’ll go halfway there. When I wrote for TV Guide and People, I supposedly had an audience northward of 20 million. I’ll hasten to say that was utter bullshit on many levels – the idea that one could trust syndicated research to count readers (as opposed to purchasers) and the presumption that every reader read every page (or ad – which is the real bubble in old media). Still, those were the numbers we bragged about, as if they gave us authority.

Dare I say that this blog gives me more authority – in Naughton’s and Lukes’ terms – than those publications did? My hackneyed example of Dell Hell reached more people in a more meaningful way than any review of Babylon 5 (though I still get in trouble for panning it).

But note well that the authority in Dell Hell was not me. I didn’t have authority (I didn’t write about PCs or pretend to any expertise in customer service). It was my message that had authority or at least relevance, as that was the reason it was passed around. And it was the passing around that invested it with authority.

So to that extent, Le Meur’s not wrong when he tries to find a way to express and calculate the idea that it’s not the author who holds authority but his or her audience. But his critics are also right when they say that number of followers won’t get him there. I think there is no easy measure, but if it exists it will be found instead in relationships: seeing how an idea spreads (because it is relevant and resonates) and what role people have in that (creating the idea, finding it, spreading it, analyzing it) and what one thinks of those people (when tells me that John Naughton follows someone, I’ll see more authority in that than, say, whom Robert Scoble follows – no offense, Robert – because Naughton is so highly selective). That is what the totality of the press-sphere will also look like as various players add varying value to add up to a whole (and in 3D, the sphere will look different to each of us, so one-size-fits-all measurements will become even more meaningless).

Part of the problem in the Twitter discussion is also that the number of followers is, in the end, a proxy for celebrity while links – which Google PageRank and, for better or worse, Technorati value – come closer to measuring at least relevance. As old media faced more and more competition it became more and more about fame (and that was when access to the celebrity became more valuable than access to the audience). The internet’s value is that it is more about relevance. So I think the reason some people reacted so much from the gut against Le Meur’s suggestion is that it unwittingly corrupted the new world with the crass celebrity of the old. The last thing we need or want in the web is Nielsen ratings.

: LATER: Case in point: Tim O’Reilly kindly retweets my link to this post and then I watch it get re-retweeted again and again. That happens because it’s O’Reilly retweeting and he has authority not becauase he has the most followers – though he has many – but because he’s smart and respected (he has authority); it also happens, perhaps, because my post is relevant to a discussion. Message + spreader (or author) comes closer to authority than mere reader ratings.

  • “… Technorati value … unwittingly corrupted the new world with the crass celebrity of the old”?

    cough – What does Technorati call that value?


  • I am just asking for a simple filter feature, should just not have used the word authority excuse my french and see you in Davos :)

  • Seth,
    Yes, that’s just what I’m saying. It’s not authority, either.

  • Pingback: My Weekend in Social Media | The Building Profits Blog()

  • Measuring online authority is difficult. After publishing a Twitter Top-40, I received a lot of criticism about measuring the amount of tweets. Those critics thought that the amount of followers was the ultimate criterium, as if it says everything. Well, is doesn’t. Of course it helps, just like the amount of tweets you produce, the amount of interaction you have with your followers and – not to forget – the authority of your followers. There is no ultimate criterium, but any relevant optional filter would be nice.

  • How close does come to measuring this? And could Neilsen not develop such a ‘standard’ tool for Internet-based media?

  • I think what Seth is feeling is… where were you a few years back? Heh.

    You’re even using the equation “Attention+Influence” which is what I used to call this a measure of – “attention influence” or “attention driving influence”.

    That word – “authority” has been mighty misused.

    Good post.

  • Anne Johnson

    Could what you are looking for be the rate of spread of meme ? _If_ (and it’s a non trivial requirement) the meme can be encapsulated in a few terms, it is possible to measure the number of occurrences of those terms, at particular times and places, by measuring what goes by on the network, which makes the measurement independent of the servers and services to which the meme is propagated.

  • Just a reminder – Clay Shirky pretty much wrote *the* piece on this:

    Clay Shirky: “Power Laws, Weblogs, and Inequality ”

  • If Loic didn’t follow so many spammers maybe his filter would not be so broken :) The problem with Twitter and numbers of followers is that it is so easy to game. Auto Follow+RT everyone+answer everyone=more followers. It’s a simple game “follow me and i’ll follow you” which of course is similar to the myspace “add me and i’ll add you” game.

    I wrote previously that I though MLM/pyramid schemes would kill twitter (mainly because they have a business model), but now i’m beginning to think that Twitter and all social media is a kind of pyramid scheme :)

    Of course the reason Twitter and other social media like facebook look and act like a pyramid scheme (it all looks great as long as it’s growing) is because they exhibit amazing network effects, and the faster growing the network the more attractive it is. The trick for any social network is to try and create some value as they grow rapidly because once the bubble stops growing it will look a lot less attractive.

  • Karl Long,
    Loic is also raising a question about search, which isn’t affected by following now.

  • Pingback: Defining the new for 2009 | AccMan()

  • Note Jeff, that I am sure that if I did a search for people who had issues with Dells before you, I would find plenty who discussed them online, in blogs and elsewhere.

    While you may not have spoke with authority on Dell issues, you certainly had attention driving and focusing influence that you did not squander when you applied it to the subject matter. You can’t discount the you in it. Just as much as you can’t discount how you composed the message.

    Boiling it down to a mere popularity contest is a massive fail – you’re right on that.

  • Jeff,
    “Followers — or subscribers or authority — is a crap metric, for the most part. According to Daniel Romero, a scientist at the Center for Applied Mathematics at Cornell University, and HP Labs, who recently sent me his team’s latest paper on the online social networks that ‘truly matter’:

    “Even when using a very weak definition of ‘friend,’ we find that Twitter users have a very small number of friends compared to the number of followers and followees they declare. (A friend here is defined as anyone who a user has directed a post, or ‘@username,’ to at least twice.) This implies the existence of two different networks: a very dense one made up of followers and followees, and a sparser and simpler network of actual friends. The latter proves to be a more influential network in driving Twitter usage since users with many actual friends tend to post more updates than users with few actual friends. On the other hand, users with many followers or followees post updates more infrequently than those with few followers or followees.”

    The authors concluded: “[S]cholars, advertisers and political activists, see online social networks as an opportunity to study the propagation of ideas, the formation of social bonds and viral marketing, among others. This view should be tempered by our findings that a link between any two people does not necessarily imply an interaction between them. As we showed in the case of Twitter, most of the links declared within Twitter were meaningless from an interaction point of view. Thus the need to find the hidden social network; the one that matters when trying to rely on word-of-mouth to spread an idea, a belief, or a trend.”

    And the fact is that Twitter is vulnerable to false assumptions that have haunted traditional media venues for years: mainly, that audience volume and value can be implied through opt-in or subscriber status. You’ve noted this above, with your reference to your past magazines. I can’t help but think of the dozens of magazines and newspapers piling up on my living room coffee table. I suppose those print publications count me among its followers, and they tout that to advertisers and target me accordingly. But the fact is they have little meaning in my life because I barely pay attention to them; I don’t recall even passively opting into most of them. (However, they do a good job of getting the fireplace roaring.)

  • Jeff, what I meant was, how can the sentences “So I think the reason some people reacted so much from the gut against Le Meur’s suggestion is that it unwittingly corrupted the new world with the crass celebrity of the old. The last thing we need or want in the web is Nielsen ratings” be true, when Technorati has been using exactly the word “authority”, offered a search based on that “authority” – FOR YEARS!

    All Le Meur did was ask for a Twitter search like A-list darling Technorati has been searching . Why the explosion?

    It CAN’T be that there is “new world” reaction, because where was that intensity of A-list reaction against Technorati?

  • I abhor popularity contests to measure authority. A great example that authority has nothing to do with how many people follow a thought is the acceptance of Einstein’s theory of relativity. Here was a tranformative idea that took two generations to get accepted.

    Another example that keeps popping up to rebuttle authority measures based on popularity is the rediscovery of ancient wisdoms about the human condition and the essence/purpose of life. All the ‘new age’ literature is rehashing ideas that were clear to thinkers in ancient Asia 3000 years ago. It shows that most people need a lot of help connecting to real intellectual authority.

    Personally, I believe that true authority is only demonstrated by truth seekers: people that keep asking the question “why” and keep peeling the onion. However, even that must be qualified as it is quickly demonstrated that great thinkers and artists can still be on the wrong side of historical arguments like Planck and Wagner were. So they exhibited great authority in some ideas and horrible misjudgement in others.

  • Shelley Powers:

    “So I’ll say this, directly and honestly, to Dave Sifry from Technorati: Dave, you are hurting us.

    The Technorati Top 100 is too much like Google in that ‘noise’ becomes equated with ‘authority’. Rather than provide a method to expose new voices, your list becomes nothing more than a way for those on top to further cement their positions. More, it can be easily manipulated with just the release of a piece of software.”

  • Seth,
    The difference, which I wasn’t explicit enough pointing out, is that Technorati, like PageRank, relies on links rather than popularity/circulation/readership/viewership. And, yes, many have complained about Technorati’s use of the word “authority.”

  • I know that Loic’s point was about search, what I was trying to get at was that he broke his own filter by following everyone back. I don’t think Loic would be in so much need of an “authoritative” search if he was actually following people who can help him filter. The specific point he was making was the ability of brands to search on comments about them based upon the number of followers a person has is somewhat valid, and and easy problem to solve. That being said I think the Motrin Moms debacle proves that your brand can be undermined by a hyperconnected niche of people even if they don’t have a huge amount lot of followers.

  • Karl,
    Motrin Moms goes precisely to the point made in the Cornell/HP Labs study on Twitter, which I referenced above. The most influential networks are often hidden.

  • Jeff, are you saying that Technorati’s ranking and search of “authority” using number-of-blogs-which-have-linked-to-this-blog-in-the-last-six-months is
    not a form of “popularity/circulation/readership/viewership”? That it does not constitute a type of “Nielsen ratings”?

    Right, many people have complained – but rarely the people at or near the top of the Technorati ranking, not like the reaction to Le Meur.

  • Ideas about power and authority are probably some of the most complex and misused / misunderstood concepts in all of social “science”. We don’t even have a clear conceptual definition of how power and authority relate to each other (although Lukes book is a good start), much less the distinctions within the concepts. Most of my own research has been concerned with questions of media and journalistic authority, so I’ve devoted a good deal of thought to this.

    Media power (the power of the media) can be defined as:

    “the multi-dimensional capacity to control the symbolic process whereby reality is produced, maintained, repaired, and transformed.”

    I’ve found that a really useful definition for media authority (especially, though not only, occupational media authority is):

    “the right to control the symbolic process whereby reality is produced, maintained, repaired, and transformed (from James Carey); the right to intervene in the course of events, to influence the actions of others and indeed create events, by means of the production and transmission of symbolic forms (from John B. Thompson).”

    The question then, is, where does that RIGHT come from?

    I’ve got some thoughts about that, too, but this is a long comment and I’ll leave it at that for now.

  • Max, I have not seen that study but I concur, there are 1,000’s of hidden networks on twitter but they are generally hidden. I know there is a significant network of jewelry sellers on twitter, some with over 4,000 followers and they are primarily using twitter as a CRM and direct marketing tool, it’s wonderful to watch :)

    One of the reasons I think these networks are hidden is because they only emerge through communication and action. Unless you are participating in a particular network you will never know the niche network exists.

  • Seth,
    I’m saying they are slightly different and I’m not even necessary judging them; they are each flawed in different ways. Linking to is clearly different from following. Not sure why this merits so much followon. It’s a relatively minor point.

  • Karl,
    While we agree that the most important, influential networks are often hidden, the HP study also proved that there are clues — in its case, with networks of people who actually interact with one another. It is not necessarily obvious, but it makes a lot of sense — but certainly requires putting aside ego and popularity. Ironically, it seems the people most obsessed with figuring out influence also are the ones most obsessed with ego and popularity, which probably makes it all the more difficult for them to grasp.

    For the record, here’s the HP labs study:


  • Seth, I think the reaction to Le Meur is based upon the simplicity of his idea.

    There’s an equating of following someone with linking to them from a public space such as a blog. The two behaviors are different.

    I don’t need better Twitter filters. But that is because I use Twitter as communication means between friends and folks I like to “follow”, like Jeff.

    But when purposely link to someone within an article or post – it means something else all together.

    Still, the level of reaction is telling – of what I am not sure.

    The fact that there has been, and continues to be so much confusion and in some cases willful ignorance by people who do wield influence is interesting to say the least. And I think why this subject matter is so compelling is it has everything to do with how information reaches people, how we become informed by one another and the world.

    I’m looking forward to your follow up Chris.

  • Hi Jeff,

    The business about followed/follower ratio being 1.0 is massively flawed for a number of reasons. The key pint on Twitter is that I don’t control who follows me and I do control who I follow.
    So what am I supposed to do when I follow about 150 people (the monkeysphere) my choice, but about 2000 people follow me – am I supposed to a) block 1850 of the 2000? or b) follow all of them and keep following any new person who follows me just for some random metric.

    Instead I have noticed that there are three kinds of people I follow –
    a) high followee high followeer – celebrities who slavishly autofollow for fear of losing followers – these people usually depend on a high follower count for their livelihood They are “celebrities”
    b) low follower, high followee – typically these people have a discerning attitude and hand pick who they follow – they are usually authorities, like Lawrence Lessig, Jay Rosen ……
    c) medium follower, medium followee – these are people I personally find interesting because something they say makes me sit up and take notice and I want to know more about how this person thinks – I am a techie and these people tend to do/say something interesting technically or are just wacky and funny.

    So a single rule about what a good ratio is just flawed on so many levels.

  • … aargh b) in my previous post should read low followee , high follower

  • Bridgett

    After reading your post, I can’t help but think of the folks that Seth Godin terms as “sneezers”. My own internal filter simply names those with lots of followers (but few people whom they follow) as the sneezers. Any business would/should go for the rapport of the sneezer. It’s not the sneezer with the info – the sneezer has the audience/tribe.

  • If I were to think about a metric to find the most influential folks on Twitter – it would have nothing to do with number of friends and have everything to do with number of re-tweets.

  • Exactly! I retweeted Tim O’Reilly’s tweet into friendfeed. That’s where the real filtering action is happening, actually, although it’s a bit hard to expose so far.

    See you in Davos. Would love to have a conversation about this there.

  • I think everyone is missing the forest for the trees. It is not about defining authority, but about creating a tool that allows users to filter data in ways that are meaningful to them. At it’s core this is a data analysis issue. A filter that works well for Loic may not be a filter that is valuable to Scoble or anyone else, but in the end it is just a method for doing data analysis to filter the wheat from the chaff. I believe the real solution would be to provide a much richer search feature that allowed me to run queries on any of the meta-data stored within twitter and to group and sort that data in a way that is meaningful/useful to me.

  • Karl – regarding “Still, the level of reaction is telling – of what I am not sure.” – I think it’s about the rawest and most blatant Mismeasure Of Man applied demonstration that I’ve ever seen. I suspect the missing piece is that on a pure-followers metric, Twitter messages from MSM reporters using their MSM accounts will tend to crowd out the A-list bloggers who are big fish in small ponds. Hence the horror, that did not happen for the Technorati top 100 and authority search, where they tended to come out on top.

  • Mark: “How close does come to measuring this {authority}? And could Neilsen not develop such a ’standard’ tool for Internet-based media?”

    We can’t the answer to either portion of this question until we agree on the meaning of ‘authority’, a very slippery and subtle idea. Authority varies by domain, medium of communication, audience, and context.

    The most intractable part of the equation is audience, because each individual in the audience applies their own set of rules to the problem of judging authority – and a good half of those rules are subconscious. Any definition we adopt necessarily has to be an intentionally multi-dimensional and fuzzy cloud.

  • @Max: I second your opinion. In my opinion actual conversational networks are the most meaningful measurement object on Twitter. Talking to another Twitter user using the @reply syntax seems to be a valid indicator for giving attention to another user. The network of people a user is replying to regularly (and perhaps receiving many replies from it as well) is relevant for this user. This is of course not authority or popularity but a third dimension – relevance.

    I also read the HP-paper with great interest. The idea of a hidden network has been so appealing to me that I had to put it into practice asap with TwitterFriends.

  • Alex Gollner

    Joe Brinkman has it right. Instead of a Web 1.0 Google PageRank method for filtering the stream or searching tweets, we should be able to use a ‘roll-your-own’ ranking system.

    A good place to start would be for each of us to have a way of turning the method we use whether to follow someone into a ‘TweetRank’ algorithm of our own. The reason I choose to not follow might be different from yours, but it’ll do for now.

    (Slightly) Further expanded on at

  • Glyn

    I don’t believe that you can say that you *still* get in trouble for criticizing Babylon 5 when the article that you link to as evidence is over four years old.

  • I often hear a criticism of the MSM that they are talking to themselves. News moved beyond them, yet they still passionately go about yesterdays news and yesterdays issues. Could this argument similarly be narrow cast. Meaning a conversation between predictable bloggers, linking to each other with no relevance to the greater culture?

  • The metrics proposed are simplistic and in any case measuring attention not authority The metrics are still based on broadcast measures (number viewers/listeners, circulation) which are potentially meaningless on the web.

    There were several mentions of PageRank (PR) in the comments. Interesting PR does not state that a page is of high relevance merely because it has many incoming links. A page will have a higher relevance if it has links coming from more relevant sites. This sounds circular and it is, which is why the PR relevance measure is computed for all pages at once
    rather than page by page. You can read about how PR works here

    In any case, authority may not be a useful filter. The issue with Twitter is that is is being promoted as a scale-free service: send as many messages as you want, follow as many people as you want, and so on. Apart from notable system outages, many people are feeling the strain.

  • Pingback: What Counts as Authority Online? « J-School: Educating Independent Journalists()

  • Pingback: AntyWeb | Archiwum » Czy mo?na mierzy? autorytet poprzez ilo?ci (obserwuj?cych lub czytaj?cych)?()

  • @Luke Yes, Twitter is scale-free. But at the same time it’s also a scale-free network ( You can follow as many people as you want and send as many messages as you want – but the messages you are sending are not evenly distributed. It looks more like a power law distribution: there are some people on Twitter you are replying to more often than others. I’m following ~1,400 Twitter users at the moment. But I’m not talking to all of them. In my last 500 Tweets, I only replied to 60 of those 1,400 more than once ( Maybe the number of incoming replies or the number of replying users would be a better measure than the number of incoming contact links?

  • @Benedikt Yes I agree that we need a network metric, a la PageRank, rather than the individual follower metric- I am sure that the HP people mentioned above can do the calculation if they have the data.

  • Pingback: Authority, Recency and Relevance are Equally Important in Social Search | Gauravonomics Blog()

  • @Benedikt you are right that I used scale-free in the wrong technical sense. I should of said something like infintely scalable meaning that the size and capacity of the twitter service can is assumed to grow without limits. Even within this framework the structure of the connection can follow a scale-free topology of preferential communication. The question here is whether that locality in that preferential structure, indicative of followers, is a proxy for authority – and I think not.

  • Pingback: My kids use to have make believe friends too()

  • Pingback: Ben Barren - Confessions of a Mad Man » a new twitter app a day keeps the wrestler not away.()

  • The Social networking websites or the Business Social networking blogs are increasingly in need of new tool to keep the great time on the web!
    It Just about time !

  • Pingback: On Twitter, Followers Aren’t Really Friends()

  • Pingback: Czy mo?na mierzy? autorytet poprzez ilo?ci (obserwuj?cych lub czytaj?cych)? | Tvn24h()

  • Pingback: Czy mo?na mierzy? autorytet poprzez ilo?ci (obserwuj?cych lub czytaj?cych)?()

  • The word “Authority” really bothers me. It smells like old style of management : the one with “command and control” rule, “command – execution”.
    We all know such conceptions belong to the past (in real life, even if some corporations still stick to it).
    In a world filled with user generated content, decentralization, where leaders aren’t any longer executives but maybe low management standouts, unofficial but really effective leaders, in such a world, authority hasn’t any longer the sense it used to have.
    Related to the topic, the number of followers isn’t of great help, the reliability of the followers (their leadership) is of. Metrics should track and translate this specific parameter.
    This is “in line” with Jeff Jarvis’ conceptions of the future of the news industry and the future of marketing, where qualified user generated content will certainly be key.

  • Pingback: Czy mo?na mierzy? autorytet poprzez ilo?ci (obserwuj?cych lub czytaj?cych)?()

  • on the right spot

  • Pingback: FutureGov » Useful links » links for 2008-12-29()

  • This might be re-purposed as an argument in favor of “authority”:
    The pipes are getting pertty clogged out there.

  • Mike G

    The number of tweets is an objective measure of quality.

    Just as the movie on YouTube with the most views is the greatest movie ever made…

  • Stephanie– in a world of user-generated content, infinite choice, and relative freedom to “consume” whatever information you want, doesn’t authority matter more, not less?

  • It’s interesting to me that this conversation seems to be connecting political theory with communications theory. We’re crossing over from how a message gets disseminated through relationships to the power of the connections in that dissemination. A lot of this post is really talking about the influential power of relationships. I would argue that political parties demonstrate strong bond connections whereas following Scoble because you’re a newbie is more like a weak bond connection. Should fascinating to watch where this goes.

  • Pingback: Tweet tweet search | Internet Marketing Blog()

  • Pingback: Twitter: disquisizioni su autorità, influenza e popolarità - Stalkk.ed()

  • Pingback: Hypercrit – Online authority debate centering on Twitter()

  • Pingback: links for 2008-12-29 « Brain Music –Gadgets, Neuroscience, Music, Social Media, Health 2.0 & More()

  • Pingback: Journalists as experts |, by Neil McIntosh()

  • Pingback: Big doesn’t necessarily mean authoritative | ()

  • Pingback: The 16 Second Attention Span of An Entrepreneur - Still With Me?()

  • Pingback: Influenceur, autorité, passeur de culture ou l’un de ces singes exubérants | Mario tout de go()

  • Pingback: The Race to Getting More Followers()