COMMENTS
interesting data. would be good to also include some analysis:
- any idea WHY ~22 tpd is the sweet spot?
- is this correlation or causation? for example - is it that the most interesting people just happen to tweet at that frequency, or is it that this is the "goldilocks" number - not too frequent, not too seldom - to be most productive at attracting followers? Or some other phenomenon?
Thank you for this report! I always wondered how many tweets that I should do on a daily basis. This is interesting.
@ilya it is only correlation, and while I haven't measured it, there must be some causation going on either TPD->Followers, the reverse or both.
Great work, guys/gals. This is great data to pass along to all of our clients. Now if we, or you, could convince them that 15-20 tweets a day only takes a few minutes total, that would help immensely. Responding and monitoring is another story ...
DW
Interesting data, thanks for collecting it. I don't think there is a number of Tweets that make a user successful on Twitter.
22 TPD comes out to roughly one message every 55 minutes (avg). I know that the people I engage most don't post that much.
However, as it pertains to the publishing -- content producers are more likely to benefit from pumping out great content hourly. NYTimes, CNN Breaking News, and TechCrunch are all feed-based Twitter account (tweeting out mostly RSS feeds).
Regardless, these are excellent stats on the higher echelons of Twitter and the frequency of updates. :)
I would love to see some stats on Re-Tweeting like the most popular RT'd items, ideal times to RT, Exposure from RT'ing, etc.
~Joe
it seems that if u tweet more than 22 times a day u might be breaking the twitters annoying limit and should expect to loose your followers. Isn't?
Love that HS analysis
Forgive my statistical ignorance, but it appears as though people who tweet approx. 40 times per day have many more followers, over 1000 on average. There's a distinct spike at that TPD rate.
Doesn't that seem to indicate that 40 is the "sweet spot", not 22? I'm sure I've just missed some obvious fact, but that's the way that initial graph looks to me.
Answer me on Twitter too, if you want, "@Mallikarjunan".
Yesch ... @DrBlockbuster finds this interesting.
HOWEVER ... @DrBlockbuster has disenfranchised (unfollowed) some of the heavy tweeters cos they fill up my page and I cannnnnie see the wood for the trees! :smiles:
Soooo ... if 22 is a good average number of tweets let's see those tweets spread round the clock and not all one after another.
@DrBlockbuster may be newish to Twitter, but he's still a very sharp, smart, kilted clansman from Braveheart country.
Slange Var
Dr Blockbuster
Networks & Forums guru
I've found myself unfollowing folks posting several times an hour or doing the 12 tweets at a time thing more than a couple times a week.
It's just too distracting with all of my other responsibilities.
I'll glance through the tweets once an hour or so, but even then if there's more 20-30, I'll clear out the old ones without reading.
I think this is really interesting
Interesting use of data; nice to have someone actually base an observation on data, rather than giving an opinion with no backup.
That said, doesn't the analysis presume that the goal of Twitter is to have the most followers?
Is that, in fact, a worthy goal?
This could launch the whole social media discussion about the value of large numbers of connections versus smaller numbers of "good" connections. LinkedIn, for example.
Twitter seems to have a lower bar of association than that of LinkedIn (except for the "LION"s who seem to think it's all about having the most connections, even if each one has no "real" connection to you.
Unfortunately, between my average number of tweets per day and my other responsibilities, all I have time to do here is to pose the question. :)
Steve, Steve re "all I have time to do here is to pose the question. :)"
... @DrBlockbuster is calling that what it is ... a cop out! :roflmAo:
Steve ... fair enough then, good fro you to recognise your guilt. No therapy required then. @DrBlockbuster is pleased :roflmAo:
...unlike this reprobate .... >>> http://tinyurl.com/d5lhxa
The information you have collected is great. 22 TPD is a good way to get more visitors
John you are drawing an invalid conclusion. This data shows tweets per day affecting followers of that tweeter, not the affect of number of followers on getting NEW visitors. If you have a static number of followers there is no one new.
We would need to see a TPD versus changeover rate of followers to draw your conclusion. That data was not presented here.
While these stats are interesting I respectively ask "so what?" The purpose of Twitter for business users is to connect with meaningful business contacts. Ask @scottmonty if he is on Twitter to make new friends or to help Ford Motor strengthen its brand?
People with large followings get them from inherent popularity or notoriety or status.
The only inluence TPD has on me is if someone Tweets too much. As othes have said, I unfollow Tweeple who clutter my TweetDeck inbox.
I'm curious, what is @karenrubin's TPD? :)
What would have made the stats more meaningful is segmentation into distinct groups rather than lumping everyone into one bucket and averaging the tweets. Averages can be very misleading. For example: Tweets by NYT should be compared to its peers in media not Barack Obama.
Interesting info. Curious to know how you know the number of days Twitters have been tweeting. Kevin Rose's average number of tweets per day seems low. Just an observation from someone who follows him.
@boumanblog
great great great and interesting
Very interesting. Means I should tweet more ?
I suggest a test : you people who tweet much less than 22 TPD, start tweeting more and look at the number of followers. Then come back to this post to add your conclusions.
These people fall into categories: politicians, celebrities, news organizations, tech types. I can see tweeting 22 + times as any member of these groups but for businesses, it is kind of an overkill. I guess it depends on what you want to accomplish with Twitter.
The data means nothing really. The most popular like Barack Obama hardly ever tweet. On top of that, it's not really him tweeting. Secondly, without any real relationships followers = Nothing. Thats the problem with everyone trying to get more followers. It means ilch unless you create real relationships. Who cares ho many followers you have? It's just an ego number.
hmmm..I have 2 Twitter accounts and Tweet about 1-5X's per week, yet have a total of about 2000 followers. I would think a bigger variable would be the product/info you are twittering about AND the other social media you are using to draw people to you i.e. inbound marketing (hellOOooo earth to hubspot:). Now having said that if you do have good product/info, tweeting about it say 22X a day would probably gain you more followers, but that is because people are interested in what you have to offer, not because you tweet a lot. If someone was tweeting about their love of cricket in the U.S., i'm guessing a thousand tweets a day won't do much good. (apologies to cricket enthusiasts).
This is a very good article....its great to see some data compiled about Twitter. Helps us who are just starting out to get an idea of what we should be doing and what kind of results we can expect to get.
http://blog.officeclip.info/2009/03/making-web-20-work-for-you/
I actually get more followers when someone else mentions me then I do when I tweet 20+ times a day. You can analyze the numbers a ton, but I try to interact and add value, and not worry too much about the metrics. I wouldn't want to tweet useless info just to hit that magic number.
I have recently started a t-shirt business and have been working in different ways to get the word out and create interest in our website and t-shirt designs. Twitter is something we have considered. I was wondering if others have seen great success with Twitter drawing people to your website for the purpose of generating sales or selling advertising on your site. It seems that several people who have posted have a large number of followers on Twitter and I would love to hear from you.
Seems to me that tweeting 20+ times per day would be distracting to your day, but if it is useful in generating traffic and sales then it may be worthwhile.
@scrossland also disengages from (unfollows), or chooses not to follow in the first place those who produce a lot of tweets. There is simply a limit to how much I can digest in between the really interesting and valuable stuff.
I'd be more interested to see the click-through rates for heavy tweeters and how many of the followers end up on the users website. In other words, for those of us who view Twitter as a means to generate web traffic, what are the stats about that?
Steve
Lots of interesting correlations here. I don't think there HAS to be a causal relationship of any kind between TPD & # of followers, or a single factor causing both, however. Seems like a red herring.
It would be interesting to know how the relationship between followers, average links per tweet, and the PageRank of the pages linked to. I would expect to find both a certain type of follower, and a large number of followers for people who link a lot to authoritative sites. Do you have these data? You guys have so many fun toys to play with!
This survey certainly had more data than my own Twitter survey (http://tinyurl.com/btd9lv) - but asks totally different questions. Thanks!
Time to start tweeting what I ate for breakfast like John Chow! (j/k)
The data and interpretation seem only to take into consideration the behavior of the Tweeter. There is another side and that is the reader. The more Tweets the more opportunity for readers to see them. I don't leave Twitter up all day watching it. So, I tend to see only a fraction of the days entire Tweets. Someone who is tweeting often and perhaps at regular intervals, will have a better chance of catching my eye. The data here is pretty linear and therefore very difficult to draw any meaningful conclusions from.
I'm with Joseph Manna, I'd like to seem more stats on things like, time of day, RT's, Tweets that have links, frequency of Tweets vs shear number, etc.
22 comments a day.. isn't that too many.. that means a user will need to have 8000+ tweets more by next year.
Twitter on
wow, that sucks. i have only 60 followers and i tweet 12 times a day.
does that mean i'm below the average of (about) 290 tweets per day for everyone else who tweets 12x/day? probably because of my tweets of drugs, tweets with words that aren't real but are invented for enthusiasm, lots of cursing, in 3 languages, have used the n-word....
shit, that sucks.
i think that equates to me not being cool... right?
So 22 TPD is the tweetspot. nice one.
How long ago were these stats gathered? @aplusk for example has over 1,000,000 followers now, which is quite larger than the 143,299 shown here. That 1 million mark sparked a huge buzz about getting more followers and I'm just curious if these much larger follower based accounts have different activity?
Very interesting. Thanks, I liked this.
Interesting! I noticed that many of the most followed also have a greater impact on the masses than the average tweeter; Obama, NYT, etc.
So is it due to the frequency of their Tweeting, or due to their overall relevance to the greater population. (Might not hold up for Britney, but hey, she was famous before, so in theory, she fits the profile, too.)
Completely inaccurate! Why did you not add @adventuregirl? She has more than 446,000 followers- is an active tweeter and retweetr and Would be interesting to see what you think of her site.
More of the #UselessTwitterMetric variety if you ask me. You seem to feel that Twitter is only for one thing, and then based on that false assumption, draw all sorts of spurious conclusions. I average right around 9 tweets per follower, meaning, if I tweet nine times, I get another person to follow me. This might be a more meaningful calculation, if you're just going by the numbers, however, there are other factors involved that have absolutely nothing to do with the raw data.
22 TPD may be optimal for some, but I average 14.6 TPD, and I find that is optimal for me.