Twitter followers don’t equal influence

A few of us have been saying this for some time, so it’s good to see that Meeyoung Cha’s research backs us up! From Scott Berinato on Harvard Business Review:

Cha called her paper, “The Million Follower Fallacy,” a term that comes from work by Adi Avnit. Avnit posited that the number of followers of a Tweeter is largely meaningless, and Cha, after looking at data from all 52 million Twitter accounts (and, more closely, at the 6 million “active users”) seems to have proven Avnit right. “Popular users who have a high indegree [number of followers] are not necessarily influential in terms of spawning retweets or mentions,” she writes.

Berinato’s interview with Cha in that post is also very interesting, and whilst some of her conclusions might just be confirming our existing gut feelings, it is very good to have some proper evidence upon which we can build.

Reading the comments to Berinato’s piece, however, leads me to think that some people are misinterpreting Cha’s conclusions. She’s not saying that social media has no use, she’s saying that follower numbers are not the right metric to measure influence (just like traffic stats for blogs don’t always correlate to their influence). The baby should not be thrown out with the bathwater.

Why I’m a fan of small talk in business

Derek Sivers reminds us that on the other end of our keyboard there lies a real person, someone who has real feelings, who will have real reactions to what we say.

When we yell at our car or coffee machine, it’s fine because they’re just mechanical appliances.

So when we yell at a website or company, using our computer or phone appliance, we forget it’s not an appliance, but a person that’s affected.

It’s dehumanizing to have thousands of people passing through our computer screens, so we do things we’d never do if they were sitting next to us.

He’s right. I’ve recently had an experience with someone suffering a total empathy failure, who didn’t seem able to put himself in my shoes and ask himself, “So, how would I feel about this situation?” It wasn’t very pleasant. This chap seemed to have entirely forgotten that their was another human being, with real feelings, who was being directly affected by his poor behaviour.

But I think we can do something about the dehumanising aspect of device-mediated interactions, and that something is to use more social media, particularly the tools that encourage small talk and phatic communication. In 2004, David Weinberger said in his JOHO newsletter:

[…] Art expresses something big in something small. (If it expresses something small in something big, you leave during the intermission.) Likewise, in small talk, we express ourselves in the details of what we talk about, the words we use, the ones we don’t, how far we lean forward, how tentatively or aggressively we probe for shared ground. Because all of this is implicitly presented, it tends to give a more accurate picture of who we are and what we care about than big, explicit conversations.

[…] I’m more of a constructivist than an archaeologist when it comes to social relationships. My aim isn’t to expose my buried self to you. It’s to build a conversation and then a relationship that eventually is so deep that we can’t disentangle the roots. For that, we need lots and lots of ambiguity.

He is still spot on. I responded to him in a post on Headshift’s blog, where I was writing at the time, and said:

What are the best aspects of conferences? The bits inbetween the panels and Q&A sessions where we get to chat with our peers. What is the best bit of the working day? Those watercooler conversations or lunch down the pub. Why do smokers have an advantage in the workplace? Because they take regular smoke breaks where they get the opportunities to chat and exchange scraps of information that become important later on.

Small talk is part of the ‘social grooming’ that is required to create and maintain social bonds. Through small talk, people reveal contextual information that they couldn’t otherwise share, particularly in a business setting. It’s around the coffee machine that you’re most likely to find out that your colleague was up all night with their sick child, which is why they looked like they were nodding off in a meeting. This extra nugget of information allows you to sympathise with them instead of getting annoyed – the context turns a negative reaction into a positive one, and helps keep the team working together instead of fostering mistrust and other destructive emotions.

Yet small talk is often despised, particularly in a work environment where one ‘should’ be concentrating on the task in hand, not chatting. But without small talk, without those bonds and the trust that they engender, teams fragment and become inefficient. The strong work ethic that has become prevalent since the industrial revolution has lessened tolerance for the social grooming activities upon which a sense of community depends, yet some companies spend a lot of money on team-building exercises which are really nothing more than formalised (and therefore often ineffective) opportunities for small talk.

The demise of the communal teabreak in offices has probably done more harm that good. The habit in many offices is that people work through their breaks, including lunch, and the idea of taking a short break mid-morning and mid-afternoon is very much frowned upon. People also have a tendency not to take breaks communally anymore except for the odd lunch or drinks after work. These trends decrease the opportunity for face-to-face small talk in the workplace.

Instead, people use email, instant messaging programme or external blogs or bulletin boards in order to get their fix of chitchat. The social requirement for small talk hasn’t gone away, it’s just moved online.

At the Social Tools for Enterprise Symposium, Euan Semple talked about his experiences implementing social software internally at the BBC. He found that a significant fraction of posts on the bulletin boards were not overtly to do with work, but either passing on experiences gained outside of work or the sort of small talk that glues communities together. But, as Euan says, “People get to trust each other through small talk, and I actively defend it against those who say it is not work related.”

It’s as true now as it was then.

Bookmarking your Twitter links in Delicious

When it comes to sharing links, I will confess that I tend to do so on Twitter these days, rather than Delicious. But now lets me do both at once. By hooking up my Twitter account to my Delicious account, I can now send a link to Twitter and have it automatically saved to Delicious. Settings let me control which links are saved, so I can specify a hashtag which will tell which of my links to save. can also convert other hashtags to tags for the bookmark saved or exclude Tweets with specified hashtags. Further settings allow relatively fine-grained control of what gets saved and how.

I’ve long since felt that Delicious is being a bit left behind. Although it’s a really useful tool that I recommend to many of my clients, it lacks the vibrant ecosystem that, say, Twitter enjoys. I’m not going to say that the development of will single-handedly change all that, but it is nice to see someone thinking about how Delicious can be worked into their day-to-day social media life.

How Twitter makes us more productive

Brendan Koerner writes over at Wired about How Twitter and Facebook Make Us More Productive. He says:

Last year, Nucleus Research warned that Facebook shaves 1.5 percent off total office productivity; a Morse survey estimated that on-the-job social networking costs British companies $2.2 billion a year.

But for knowledge workers charged with transforming ideas into products — whether gadgets, code, or even Wired articles — goofing off isn’t the enemy. In fact, regularly stepping back from the project at hand can be essential to success. And social networks are particularly well suited to stoking the creative mind.

Brendan makes the point that surveys like Nucleus Research’s or Morse’s, assume that all Twitter/Facebook activity is wasted, but in reality it is not. He then goes on to discuss the human creative process, highlighting the “need periodic breaks to relieve our conscious minds of the pressure to perform — pressure that can lock us into a single mode of thinking.”

Regular breaks, it turns out, are important for our brains to process information and the “conceptual collisions” that occur when we see nuggets of unrelated information can prompt us to make mental connections that we otherwise would not have. Twitter and Facebook are, of course, great at exposing us to unexpected information.

I’d add two more points to explain why Twitter, used well, isn’t a de facto waste of time:

Firstly, Twitter is amenable to sporadic checking, which means that users can check Twitter in otherwise dead moments, e.g. waiting for a web page to load, a file to save or a phone to be answered. Quite often I check Twitter whilst I’m waiting for my computer to do something else. What else would I do with that time? Stare at my screen and wait. So net win on the time saving there.

Secondly, Twitter saves me time by connecting me to people who have answers to my questions, including some questions I didn’t know I needed to ask. I get a lot of ideas for blog posts from links that my friends post to Twitter, for example. I also often get my answers from Twitter faster than Google can manage and those answers are often higher quality and contain insight Google just can’t provide.

These productivity research companies really do need to get a clue when it comes to Twitter and produce something a bit more nuanced and less scaremongery!

How important is Twitter to your blog’s traffic stats?

Last Friday I wrote a blog post on my own blog about The Impenetrable Layer of Suck and did what I usually do with blog posts these days: I Tweeted it. I saw a few people reTweet it, so thought I’d check my stats. This is what I saw:

How important is Twitter?

I’ve heard many a time from friends at Guardian Technology, who all regularly Tweet links to new articles and blog posts, that Twitter is a greater driver of traffic than Google News. I’ve found it to be true here as well. On days that I Tweet a link, traffic is much, much higher than days I don’t.

I rarely see links from other websites listed in my referral stats, apart from my own site where there’s a feed in the sidebar and weekly roundups. The decline of the trackback is an interesting, and sad, thing. They got so polluted by spammers that they became unworkable for most people and now I rarely see functioning trackbacks. Blogrolls have also fallen into disfavour, probably because they were such a pain to keep up to date and the technology to look after your blogroll didn’t develop much functionality beyond very basic add/delete/sort links.

This is a shame. In the early days of blogging, I felt like I really was a part of this huge network of bloggers, all passionate about the opportunities this new technology gave us, all excited about the democratisation of publishing. Now blogs feel much more isolated from each other, less connected, less like the whole is greater than the sum of the parts. More like lone voices howling in the storm.

Twitter brings traffic, as sometimes does Facebook, but it doesn’t make me feel that this blog is connected into a wider network. Whilst information flows through my network, just as it did before, that flow is mostly invisible. Twitter doesn’t show me whose Tweet is sending me the traffic, it’s all just a nameless wall of The network has slipped behind a veil.

It’s great that Twitter brings readers, but I miss that sense of connection that my referrals stats used to bring me.

So, how important is Twitter to you, compared to other sources of traffic? Do you get most of your referrals from Twitter? Is Twitter now where you find most of your news?

Asymmetry: The problem with social networks

Clive Thompson writes on his blog (and in Wired) about how social networks such as Twitter become dysfunctional when the network gets too big and, as a result, too lopsided:

When you go from having a few hundred Twitter followers to ten thousand, something unexpected happens: Social networking starts to break down.

This is a point I’ve been making for a long time, not really from personal experience but from observing various friends who have very high follower counts. Clive goes on:

Technically speaking, online social-networking tools ought to be great at fostering these sorts of clusters. Blogs and Twitter and Facebook are, as Internet guru John Battelle puts it, “conversational media.” But when the conversation gets big enough, it shuts down. Not only do audiences feel estranged, the participants also start self-censoring. People who suddenly find themselves with really huge audiences often start writing more cautiously, like politicians.

When it comes to microfame, the worst place to be is in the middle of the pack. If someone’s got 1.5 million followers on Twitter, they’re one of the rare and straightforwardly famous folks online. Like a digital Oprah, they enjoy a massive audience that might even generate revenue. There’s no pretense of intimacy with their audience, so there’s no conversation to spoil. Meanwhile, if you have a hundred followers, you’re clearly just chatting with pals. It’s the middle ground — when someone amasses, say, tens of thousands of followers — where the social contract of social media becomes murky.

‘Microfame’ is a term I first heard used by Danny O’Brien at OpenTech 2005 in his keynote, Living Life in Public (available from the UKUUG site, embed coming soon!).

This was in an era before Twitter, before Facebook opened up to the world, when most people became ‘internet famous’ through their blog. But becoming ‘microfamous’ puts people at the centre of an uncomfortable social dynamic. As Danny said:

There are people out there who know something about you, but you have relatively little knowledge about them.

This becomes problematic because the microfamous rarely have the resources that the truly famous do to protect their privacy. But more importantly, it creates a disconnect, an unbalanced power relationship that we don’t really have the societal experience to understand. Knowledge is, after all, power.

This relationship asymmetry has been amplified by Twitter especially. Twitter is a very good example of how poorly we understand these dynamics and how the tools that we create and use are not designed to take the microfame effect into account.

It’s appears that there are a number of stages in the growing asymmetry of one’s Twitter network. The first is when the majority of @ messages you receive come from people you don’t know. That happened a while ago for me, probably at around the 2000 follower mark. Then @ messages from people you know get swamped by @ messages from people you don’t. Finally, the @ messages to every last thing you say flood in, killing your ability to have a conversation with anyone and making it impossible to build connections.

I’ve not experienced those last two stages, but I’ve seen it happen to friends and it’s not pretty. It puts them in a difficult position where the people @ing them feel put out that they don’t get a personal reply, but the amount of time it would take to read and respond to every @ makes it extremely difficult.

This is the eternal problem of social networks. In order to be financially successful, social networks need to grow large. But in order to be socially successful, they need to stay small. Seemic was a good example of this. In the early days, it felt like a small, intimate community where one could upload a video and have a real conversation around it. As it grew, the conversational seeds, those first video uploads that broached a new subject, became so numerous that it was hard to find one’s own, let alone the responses to it. In fact, it became so time-consuming to participate I had to give up.

With Twitter, the problem is just as much about the tools as the network itself. Twitter clients tend to be designed for people with small networks and don’t deal well with asymmetry. Most tools, for example, have two ways to show @ messages: you can see @s from your friends in your timeline or see all @ messages lumped together, regardless of who they came from.

I’ve yet to see a tool (although clearly I’ve not used all Twitter clients) that gave you a third choice, to see all @s from people that you follow in a separate view. That would at least allow the Twitterer to focus on maintaining relationships with the people they have chosen to follow, whilst facilitating a dip into the faster-flowing stream of @s from the rest of Twitter whenever they wanted.

It might be tempting to dismiss this problem as one that only the cool kids suffer from, but that would be to miss the wider point. In some situations, creating small trusted networks with variably-permeable boundaries is key to creating a sustainable broader network. This is particularly of collaboration spaces, where you want to invite only key people to work with you, although that group may change from project to project.

(Now, you may think that Facebook achieves this, but it doesn’t. It gives one the sense of being in a small sub-community without actually delivering on that promise – the boundaries are far too porous, and their porosity is not entirely under your control.)

We need to do a lot more thinking about this problem. It’s relevant in a whole host of context – hot-desking enterprise, for example – and most social networks focus on creating broad opportunities for interaction without considering how to let people create natural boundaries where they feel comfortable.

David Carr on Twitter

There’s a lot of stuff written about Twitter and most of it rubbish, but every now and again I read something that really sums Twitter up nicely. This piece by David Carr in the New York Times is one of those great articles that talks very clearly about why Twitter is both useful and important, but without flipping out in gushing hyperbole. It’s the sort of thing that I’ll keep in my arsenal of articles to show people who want to understand social media.

Twitter announces bylines

Twitter is beta testing some functionality specifically for business users: the byline. If you’re running a business account you might want more than one person to be able to use it, but it can cause confusion to have more than one person Tweeting under a single identity. CoTweet handles this by allowing multiple people to access a single account, with each person specifying initial to be automatically appended to their Tweets.

Twitter is taking this idea to its logical conclusion by providing a proper byline:


This is a neat bit of functionality and its inclusion in the API should make life easier for CoTweet and Hootsuite users (not to mention other clients that will be able to incorporate it). Twitter is testing it at the moment amongst a limited group of business users ahead of a full launch to “all business users and ecosystem partners”.

Google’s real-time search ups the misTweet ante

Google has announced that it is going to be indexing the web in real time:

Now, immediately after conducting a search, you can see live updates from people on popular sites like Twitter and FriendFeed, as well as headlines from news and blog posts published just seconds before.

[…] You can also filter your results to see only “Updates” from micro-blogs like Twitter, FriendFeed, Jaiku and others.

[…] Our real-time search features are based on more than a dozen new search technologies that enable us to monitor more than a billion documents and process hundreds of millions of real-time changes each day. Of course, none of this would be possible without the support of our new partners that we’re announcing today: Facebook, MySpace, FriendFeed, Jaiku and — along with Twitter, which we announced a few weeks ago.

This announcement should make people with twitchy Twitter fingers pause. There was once a time when a mis-posted Tweet could be deleted in time to ensure it never made it into Google’s cache (although never fast enough ensure no one saw it in their timeline). Google hasn’t explained how they will now deal with deleted updates, but my own experiment this morning showed that deleted Tweets are not deleted from Google in a timely fashion (if at all).

This is good and bad news. On the one hand, Google Cache has allowed me to do a bit of forensic Twitter searching to piece together deleted conversations. There will be times when it will be an important tool for holding public figures accountable for what they say in public. On the other hand, everyone makes mistakes. Shouldn’t we be able to delete and forget them?

However Google ultimately decides to deal with deleted content, it’s a timely reminder not to update in haste.

The role of dopamine in social media

What is it that makes our inbox such an enticing place that we spend hours there every day? It’s a question that fascinates me, mainly because I have such an uncomfortable relationship with email. I get lots of it, am often slow to respond and frequently end up feeling guilty because my email has got the best of me.

Psychologist Susan Weinschenk puts the blame for our obsession on dopamine:

[T]he latest research shows that dopamine causes seeking behavior. Dopamine causes us to want, desire, seek out, and search.


It’s not just about physical needs such as food, or sex, but also about abstract concepts. Dopamine makes us curious about ideas and fuels our searching for information. The latest research shows that it is the opoid system (separate from dopamine) that makes us feel pleasure.

Wanting vs. liking – According to Kent Berridge, these two systems, the “wanting” (dopamine) and the “liking” (opoid) are complementary. The wanting system propels us to action and the liking system makes us feel satisfied and therefore pause our seeking. If our seeking isn’t turned off at least for a little while, then we start to run in an endless loop. The latest research shows that the dopamine system is stronger than the opoid system. We seek more than we are satisfied (back to evolution… seeking is more likely to keep us alive than sitting around in a satisfied stupor).

A dopamine induced loop – With the internet, twitter, and texting we now have almost instant gratification of our desire to seek. Want to talk to someone right away? Send a text and they respond in a few seconds. Want to look up some information? Just type it into google. What to see what your friends are up to? Go to twitter or facebook. We get into a dopamine induced loop… dopamine starts us seeking, then we get rewarded for the seeking which makes us seek more. It becomes harder and harder to stop looking at email, stop texting, stop checking our cell phones to see if we have a message or a new text.

This sheds much needed light on why we spend so much time checking for new email only to then not deal with it when it has arrived, but there is more to the email problem than dopamine.

There are cultural problems around the use of email as a proxy for productivity; huge email loads being worn as a badge of honour by people who like to equate their inbox martyrdom with a commitment to work; and defensive emailing by people who feel so scared or insecure that they CC everyone. These issues around the sending of mail need to be tackled, probably before we try to tackle our dopamine-fueled inbox obsession.

But as Weinschenk points out, tools like Twitter are just as likely to “send our dopamine system raging”.

So if social media is as addictive as email, isn’t it pointless to try to replace one with the other? I don’t think so, no, because there’s more to it than trying to reduce inbox faffing, as important as that is. It’s also about improving sharing, findability, archiving, collaboration, conversation, staff relationships, morale and efficiency. These benefits, in my opinion, outweigh the potential flaws in the new tools.

We do need to be aware that social media isn’t without its problems, but understanding the fundamental biological and psychological processes that shape the way we interact with technology will help us to solve those problems. I look forward to watching and maybe even participating in the emerging field of technopsychology.