The White Noise That Is Twitter

Posted by

Don’t get too caught up in the amount of followers that someone has on Twitter or how big the overall network actually is.

Unlike some of the other new media channels, it’s important to remember that Twitter is still nascent, confusing, busy and hasn’t fully matured to the point where it has a clear direction or intention. The more people connect on Twitter, the noisier it gets. The noisier it gets, the harder it is to uncover the gems, find the most relevant conversations, and the best people to follow.

The Real-Time Web only complicates Twitter further.

Even if you have a significant amount of followers and those people are able to amplify your message, Twitter lives in the real-time Web. Meaning: if you’re tweeting while your key followers aren’t on Twitter/paying attention, your content remains stagnant in the ever-flowing river of tweets that flow in an unrelenting tsunami of 140 characters. This is why many of the more "influential" people on Twitter actually retweet their own content several times a day, across multiple time zones. They are – in essence – trying to ensure that their tweets rise above the noise.

We all can’t be Ashton Kutcher (apparently, this is a good thing).

Ashton Kutcher has close to six million followers on Twitter (I know, that’s incredible. And no, it’s not a typo), but Mashable had a recent Blog post titled, Ashton Kutcher Has Little Twitter Influence. According to the Blog post: "A study conducted at Northwestern University determined that celebrities like Ashton Kutcher with millions of Twitter followers are mostly ignored on the social media site, resulting in very little if any influence. When the researchers applied their mathematical algorithm to the countless tweets that appear on Twitter each day, they found that experts in certain fields were much more likely to cause topics of discussion to become trends. That might come as a relief to social media enthusiasts who crave discussions of substance, and a surprise to critics who argue that social media is prone to inanity." Once again, we’re looking at the whole "quality over quantity" debate.

Is there really a lot of conversation happening on Twitter?

Social Media monitoring company, Sysomos, released a study yesterday titled, Replies and Retweets on Twitter. There is no denying that a retweet is a powerful indication that what someone says on Twitter has relevance (more on that here: The Retweet Is One Of The Best Measurements Of A Brand’s Success), but the stats from this research (which studied 1.2 billion tweets in the last two months) may surprise you…

  • 29% of all tweets produced a reaction (a reply or a retweet).
  • Of that group of tweets, 19.3% were retweets and the rest replies.
  • Out of the 1.2 billion tweets only 6% were retweets.
  • 92.4% of all retweets happen within the first hour of the original tweet being published.
  • 1.63% of retweets happen in the second hour.
  • 0.94% of retweets happen in the third hour.
  • 96.9% of @ replies happen within the first hour of the original tweet being published.
  • 0.88% of replies happen in the second hour.

Beyond the retweets and replies is there a conversation?

Here’s what the Sysomos research says: "We also examined the distance between an original tweet and the replies it attracts. Of all tweets that generated a reply, 85% have only one reply. Another 10.7% attracted a reply to the original reply – the conversation was two levels deep. Only 1.53% of Twitter conversations are three levels deep – after the original tweet, there is a reply, reply to the reply, and reply to the reply of reply."

Twitter isn’t bad… it’s just different.

Yes, there are millions and millions of people on Twitter, but don’t confuse Twitter for a traditional mass media outlet. It’s not. It’s a live organism that can be one thing in this moment and something completely different in the next. You could have all of your followers active and engaged in one moment and silence in the next (depending on if they’re online or not). Twitter continues to be many different things to many different people. And, because of the many uses, people, applications, broadcasting and communications, we’re not seeing a ton of depth or amplification (as we may have once thought). That doesn’t make Twitter bad, useless or on its way to irrelevancy, it just makes Twitter something different that we’re all going to have watch and figure out.

Is Twitter nothing but white noise? What’s your take?


  1. While there is indeed a lot of noise in the Twittersphere, there are an incredible amount of real conversations happening between peers. For example, the Web 2.0 community in Government has developed an impressive Twitter presence. We can poll each other, cross-pollenate, beg each other for studies/ templates/ resources… and with the transparency of being in the public eye. It’s incredibly fast and engaged.
    I get that companies are trying to figure out how to break into Twitter as another channel. I find that the best B:C experiences are the ones where I can get info or ask for help and get a response via Twitter. Seesmic updated their Twitter platform last week and I couldn’t figure out a setting. I @’d them and they replied within minutes. Granted they ARE a microblogging company, but other tech-savvy businesses like ThinkGeek and Threadless Ts all do the same. It’s less about asking me to engage, but being responsive when I try to engage with the company. A positive response likely means that I will be willing to engage if the company asks me to in the future.
    Thanks for another insightful post 🙂

  2. Twitter is a communications tool that people use in thousands of ways. Some use it to make white noise, others use it to tweet the location of earthquake survivors. One thing for sure: it has gone mainstream and is not going away any time soon. One thing I see is that the rise of Twitter creates a real opportunity for people creating good, long form content like blogs and books because most folks are stopping at 140 characters.

  3. It’s going to be very interesting to see how big brands respond to the reality of the constant flow and small numbers. The sum is always greater than its parts, but let’s face it: brands like to know that they’re hitting a whole whack of people and that those people are helping to spread their ideas. Things are now different.

  4. …and that’s the main point: because Twitter is anything and everything, the cumulative effect can be perceived as white noise. Let’s see which brands can play with that white noise and make it work for them.

  5. Very interesting stats and ones I’ve wanted to know about RTs. I also wonder if there are studies done in regards to conversations on Twitter and if there’s a correlation to see if people who have met face-to-face then have more Twitter conversations, RT each other’s tweets more, reply, etc. I guess through my own observations I’ve noticed those face-to-face meetings have resulted in more interaction online. Would be interesting to know exactly what is the type of conversation that has a reply, to a reply, to a reply.

  6. What worries me is that some people are going to look at these statistics and decide they need to auto-tweet the same message hour after hour in hopes someone will see it and even further clog the pipeline.
    I recently unfollowed someone who did that. Suddenly it’s a lot easier to see messages (comments, links, etc.) that can make a difference in the way I serve clients. Did the person I unfollowed ever have valuable content? Actually, yes. But in what I suspect was his tactic for dealing with the tweeting at the “wrong time” problem, he lost someone who might have someday been a client or helped him be more successful.
    Lord help us if more people see those percentages and decide more repeat Tweets are the answer.

  7. In some ways twitter is like a very large version of Speaker’s Corner at Hyde Park. If you stand on a soap box and shout about your product people will quickly move on to a more interesting speaker. You either have to have a very interesting message about your product or more interesting to me hang about in the crowd talking to people and subtly push out your viewpoint. Most importantly you have to have opinions which is anathema to many companies.

  8. I like twitter, but I prefer the conversation options on Google Buzz and Facebook. I find it very difficult to have or follow a conversation. Sometimes I want to reply to what someone said, but they have like a hundred tweets after the one I found interesting. On Buzz and Facebook, the replies follow the post and you get notified by email, which to me is great. Recently, I had a conversation by DM because it was so hard to follow.

  9. Hi Mitch,
    I love it when a blog post pulls together different articles for support, it shows how deep you really dig into the subject 🙂
    I was just reading Nick Emmel’s post about over-participation and agencies’ absolute need to see numbers (post is here –> when those numbers don’t necessarily correlate to the business’s goals. A number of digital marketers are focused on just the numbers, but we’ve had clients that have monitored online conversations and seen that at that moment it hasn’t been in their best interest to directly engage with people or encourage participation for a particular project.
    Great to have your thoughts and all of the comments underneath 🙂
    Michelle @Synthesio

  10. Two things about these numbers: First, to me they highlight just how much we don’t know about Twitter usage from simply mining-for-tweets. The datastream alone simply can’t tell you enough about how Twitter is used, but because data mining is free/cheap, we all do it and try to rationalize the process ex post facto. I wrote more on what we don’t know about Twitter from this data here: If A Tweet Falls In The Forest.
    Second, the headlines on this study currently read “71% of Tweets Are Ignored” or something similar. As a researcher, I’d write this headline as “Nearly Three In Ten Tweets Provoke Reaction.” I think the fact that 29% of tweets generate at least 1 retweet or reply is an enormous number. I follow 3000 people. If, on average, each tweeted once per week (I’m being conservative, but maybe only slightly), then statistically I would be retweeting or replying to almost 900 tweets per week – well over a 100 a day. I certainly don’t do that!

  11. The very concept of white noise on Twitter always fascinates me. As I often say, Twitter is like a room full of people, most talking in little groups about their specific topic, some talking alone about their daily routine against a wall, some talking one-by-one about random things, and so on. If you switch from your group to another, you will most likely consider the new topic as “white noise”. If you approach someone talking alone, you might very well wonder why is he doing that at all.
    Everything is relative, if I join Twitter to follow Ashton Kutcher maybe I will be inclined to think Mitch Joel’s messages as useless white noise.
    In the end I think Twitter is not a single communication platform, but many platforms all working together to deliver messages which happen to be all completely public.

  12. Hi Mitch,
    From the beginning, I felt Twitter could help to improve this new distributed media many of us call social media, or it could devolve into yet another channel for mass media. I must admit I’m seeing both, but the later risks completely obscuring what is good about Twitter. Twitter’s broad success is making it the supermarket tabloid of social media – available to everyone, but with little compelling information.
    For many of us, Twitter has become a slightly more social RSS feed. I some cases it’s even powered by an RSS feed or application. The pressures to make Twitter profitable and to increase the number of users, will end up making Twitter something far less than it had the potential of being. I think we’ll be looking back at Twitter in 2 or 3 years and remembering it fondly for what it could have been!

  13. It’s precisely because of these reasons you outline—Twitter’s difference, chiefly—that make it so wonderful. I’m constantly surprised by the element of serendipity it also presents. Alas, it’s not for everyone or every business, but I find Twitter indispensible.

  14. I only have one point of debate. Otherwise, I’m with you on this. The point is your parenthetical remark “as we may have once thought” in the 2nd-to-last sentence of that last full paragraph.
    Twitter may not have been around for long, but it has certainly changed significantly and quickly, right? For me, the depth of conversation as a percentage of my total tweets has surely waned over time, so while the study’s findings are spot on at the moment, I believe it was a different environment on Twitter a year or 2 ago. Nonetheless, many of us can clearly point out current and frequent exceptions to the study’s points about depth of conversation. I would be willing to bet that the low percentage of 3, 4, 5 level deep conversations are primarily due to the mass of people, who joined Twitter in the last year only to follow celebrities, news, TMZ, etc. Those people mostly follow, broadcast, and/or respond only to massive accounts that don’t have the time to converse.
    Personally, I pay the most attention to response rates. For every tweet, @ reply, retweet, or DM I send, how many @ replies, retweets, DMs, or clicks am I (or my clients) getting. Then of course, you can go further down the line and do your CTA, conversion, CPO, etc percentages if you like. But, when sticking only to Twitter, trended response rates are, in my mind, so much better of an indication of the attention that is being paid to the work we put into this medium.
    Now that I think about it, testing what level of conversation (replying to every @ reply, only the ones that seem relevant, none, etc) a Twitter account should have is not that different from you testing out replying to every comment on your blog, right?

  15. I’d be curious to know how many tweets drive conversation beyond twitter. For instance, I read Mitch’s tweet and clicked through to read this post, to which I am now replying here — and not on twitter. I did not RT Mitch’s message (yet) and I did not reply on twitter, so for the sake of the above research, the message had little impact on me. And yet here i am. While I certainly count as a “clickthrough”, the conversation aspect is even more valuable.

  16. To mind mind — Twitter is a transit point more than anything. It is a place to get together, find something interesting, and then move to that point. Twitter > BBC, or Twitter > SearchEngineLand, or Twitter > Anywhere else.

  17. I agree!! My response to that study — which I tweeted about, natch, was “I don’t care about how many people read my tweets, as much as I care about the effect on those that do.”
    The other thing I said (which sounds similar to your “If a Tweet falls in the Forest”), is this – “if you go into a mall & talk to one person at the mall, have you really failed to talk to the thousands of others that are there?”
    The power of Twitter — is that it is BOTH an individual conversational tool AND a mass broadcasting too. The people (or brands, or groups, or whoever) who “win” on Twitter are those who know how to go back and forth between conversations and broadcasting in an interesting and compelling manner.
    Oh — and my conversation on Twitter about THIS topic included discussions with people I met (yes, on Twitter) in Turkey and Hong Kong, and then we then took it OFF Twitter to include readers of the Atlantic and other tech blogs along with Malcolm Gladwell’s piece. Just because you can’t track in on Twitter doesn’t mean it’s not happening.
    I ignore the white noise. I act on the great ideas. *that* is the power of Twitter.

  18. The value of Twitter is different for everyone (a fact borne out by the comments on this post). Some use it to spam people with inane messages, some (as Tim Vickery and Ann Kingman have pointed out above) use it as the Grand Central of the internet, others use it for let’s say “less than savoury” reasons, still others use it as a way of simple communication and conversation…
    All of the above happens every day on Twitter.
    Back to the headline of this post: Is it white noise? Yes…but only the parts you that aren’t relevant to YOU. Like falling off a log, there is no one correct way to use/consume/interact with Twitter.
    If your approach gives you the result you want, that’s all you need to know.
    PS: as for the Ashton Kutcher stat, it should be no surprise given that Kutcher’s status was not built on Twitter or social media. Kutcher is not and has never been the poster-child for what Twitter or indeed social-media is all about. I wrote a piece on it in May, if anyone’s interested (

  19. This is a tough post to comment on! There’s so many thoughts running through me head on this one. Twitter is noisy. It’s made to be that way. I think it’s extremely useful if you can find a valuable audience, which is easier to say than do. It’s definitely a challenge to find quality conversation. Twitter is not a popularity contest, or it shouldn’t be anyway.

  20. To be followed is meaningless… but to be listed? I’m starting to believe that your follower-to-listed ratio is a good metric for how well connected you are with your followers.
    To be listed means you have made an impression on someone or are associated with a particular topic in their mind. Less noise, more signal, right?
    If that’s the case, then you’re the master Mitch. You follower-to-listed ratio (FLR) of 11.87% beats out everyone from Chris Brogan (9.78%) to Avinash (10.61%) to me (8.81%). 🙂
    What do you think? Any value in such a metric?

  21. I wonder how much of the results in Sysomos’ study have to do with the fact that so many people use Twitter as a broadcast medium (the slightly more social RSS feed Jose Leal refers to above or perhaps a digital scrolling message board) with little or no intent to interact with followers.
    I also believe it has a lot to do with what Twitter is being used for. If it’s being used as a customer service vehicle then I don’t think the ‘white noise’ label applies.

  22. I doubt you would find any major percentage of people making the connections happen in real life. My guess is that it would be a tiny fraction of the Twitter population. As for conversation, wouldn’t you agree that both the retweet and the @ replies are highly indicative of the level of conversation?

  23. We have to remember that the platform was created without the infamous retweet and @ replies. The creation of those actions was a function of the community. One could easily argue that Twitter wasn’t created for conversation at this scale.

  24. It’s important that Marketers not work in absolutes. There is tons of value in listening and being reactive without necessarily being engaged. In the end, it’s near impossible to get everybody to see/believe in your own brand perspective… and there’s nothing wrong with that.

  25. The data can sometimes be fuzzy – especially when people editorialize on it (like I am doing) or when the producer of the data wants to “sell” the idea (i.e linkbait).
    Both you and Lisa raise a more than valid argument. The challenge is that its not what a brand wants to hear. They have a traditional mass media mindset. They want to know that if they go on Twitter what percentage of people will follow them and amplify the message. I know, it’s not the thing many people want to hear, but we have to be realistic about what their needs truly are.

  26. The power with Twitter is that you can ignore the sources of content you do not like. In essence, you don’t have to “see” anything you do not want to see. I like to say that Twitter is a “spam free” environment because I can block or ignore the elements that I do not like.

  27. As someone above said, twitter can be a slightly more social RSS. My point was that calling a message that does not receive a retweet or an @ reply “white noise” assumes that the relevant actions are only those within the twitter ecosystem.

  28. It’s not the individual tweets that cause the white noise. It’s the cumulative effect. That “river of news” that constantly grows and flows. From a news and media standpoint, I love the reference to Twitter as a Social RSS (or, as Chris Brogan calls it, a “serendipity engine”).

  29. Twitter is not nearly as effective for me as it used to be, both from a marketing and a conversation standpoint. There’s lots of noise; very little signal.
    I’ve been toying with the idea of developing a strong presence in places like Sprouter, almost in the same way I did with Twitter 2-3 years ago.

  30. Mitch,
    PHENOMENAL connection to this blog post in your latest podcast with Joseph Jaffe. As much as I enjoyed this post, having heard the podcast after really drove it home.
    Thanks for that.

  31. Great seeing you in person yesterday, Joe. The podcast with Jaffe will be published here at some point tomorrow… it came out great. You should know that the Blog post was published prior to us having that conversation.

  32. The way I keep Twitter relevant is both filters and private lists. This way, I am able to navigate the many connections I have and focus on the ones who are constantly and consistently adding value.

  33. Part of the stigma is that if you’re only following people that are interesting to you but you also have many people interested in following you, then you are considered a snob for not following back. I never subscribed to that… so I guess I am a Twitter snob!

  34. Funny enough, Avinash was just saying the other day that being listed is a good metric. I wonder how many people create lists? I hardly do it (seems like a lot of work ;). That being said, I would agree that being listed is definitely some kind of a thumb’s up from your community.

  35. If that’s what people are using it for… then that’s what it is 😉 As for the white noise: that doesn’t speak to quality just volume and persistence. Twitter is white noise… from the raw numbers and tweets.

  36. I wouldn’t doubt Avinash would say something like that. I heard a podcast last summer in which he mentioned the # of retweets per thousand followers per month as a metric. Blew my mind.

  37. Hi Mitch,
    I’ve been reading Clay Shirky’s Here Come Everybody and – talk about serendipity -after your post on Twitter being ‘white noise’, I came across this quote in his book: “Even in a medium that allowed for perfect interactivity for all participants (something we have a reasonable approximation of today), the limits of human cognition will mean that scale alone will kill conversation”.

  38. Thanks for writing this Joel.
    I think too many people were quick to jump on the data and say “Well, twitter isn’t working the way we thought.”
    I agree with you that many people use Twitter in different ways and it’s still so new that no one has it pinned down to a science. The thing that most people seem to forget is that just because a tweet didn’t gain a reaction (an @ reply back or a RT) does not mean that the tweet wasn’t read or the link in the tweet wasn’t clicked.
    I actually plan to have my own post about people’s reactions to the study up in the next day or so, but as usual, you beat me to the punch.
    Sheldon, community manager for Sysomos

  39. It would be interesting to see if you could tell if a tweet was “consumed” – whatever that means: read, looked at, whatever. That being said, if someone tweets and it doesn’t get an @ reply or a retweet I would surmise that it is an indication that the tweet did not resonate. That is unless there is a link embedded in the tweet (in which case, it can be measured by whether or not someone clicked).

  40. Assessing the value of a connection based on number of followers is like judging the quality of a house by counting the nails. I use twitter but block anyone who tries to sell me something or has objectionable commentary.
    When I want to learn what thought leaders and my real connections have to say, I use Digg and Gist. I have control over what I see and there is far more depth in the coverage. Because the data isn’t random, I get the quality I need.

  41. I get about 2400 tweets/day on average and I respond to maybe 30? I only see maybe 200. So that 26% sounds high but then it just means one person is reacting. I could have 1000 followers and if only 1 retweets or replies it is included in the 26% yet the response rate is only 0.1%

  42. If you’re looking for Twitter filters on the go, you should check out our TweetAgora app. It lets you mute people, keywords, conversations, and annoying services like foursquare.
    It’s also a great aggregator: you can create ‘Agoras’ to bring any number of keywords, people, and Lists into one ordered stream. Our latest version also introduced the ability to view tweets by relevance, see personal trends, and check replies on any tweet (all powered by the Cadmus API).
    It’s free to try on the App Store, so check it out!

Comments are closed.