Over the course of the last few years of presenting data about Twitter, I've been asked a few common questions that, at the time, I didn't have the data to answer. One of the most common questions asked has been regarding click-through-rates (CTR).
I started by analyzing the length of 200,000 link-containing tweets as well as the CTR the links in those tweets generated. I calculated CTR as the number of clicks on a link divided by the number of followers the user had when he/she tweeted it.
What I found was a little surprising. Up to about 130 characters, as the length of the tweet increased, so did the CTR.
The takeaway here is pretty straightforward. If you want more clicks on the content you post to Twitter, use longer tweets, but don't surpass 130 characters.
So it seems as though Twitter users are more likely to click on links in tweets if they're accompanied by something else, such as a description of the link. When publishing content to Twitter, choose your characters wisely, and use that space to encourage others to click the links you post by including an enticing description or teaser of the content you're linking to.
Did this data surprise you? Why do you think Twitter users are more likely to click on longer tweets?
Ranae 11:12 AM on September 27, 2011
Hi Dan. This data did surprise me! I would have thought people today would prefer short, concise, to the point information and did not have time to read the longer tweets. Thanks for sharing the great info!
Kimberly Peek 11:20 AM on September 27, 2011
I always love seeing the new research and graphs. To me, this one seems intuitive. People are busy and time starved. They scan information. It makes sense they would only spend the time to click through if they felt that what is behind the click will be worthwhile. In this case, the text accompanying the link has the same power as an email subject line to make or break the CTR or open rate.
cakapniaga 11:28 AM on September 27, 2011
Sometimes I used LINKBAITING method to tweet. After reading your article, I think better I change it, to maximize the allowed 130 characters.
Glenn 11:35 AM on September 27, 2011
Fascinating. Between 55 and 130 characters you have a strong R-squared for this. If you include all the data -- I'd be curious what kind of correlation you would actually have...
Pete Austin 11:37 AM on September 27, 2011
Correlation is not causation.
The evidence is also consistent with another theory: that shorter tweets are retweeted less. This is totally unsurprising because a lot of shorter tweets are not intended to be retweeted. For example:
@someone LOL
Increasing the length of your tweet will not make retweeting more likely. For example:
@someone LOL!!1!!
Ray Wright 11:38 AM on September 27, 2011
Dan, great analysis as usual. Thanks for sharing it. I'd like to follow up on Kimberly's comment. It doesn't seem too surprising because the more characters, the more words (most likely), the more terms to search on and find.
Also I was puzzled as to why you would divide the number of clicks by the number of followers? Tweets are available to everyone and will certainly be noticed by more than followers if they contain hash tags and popular "keywords" (not sure if that's the right term for twitter search terms?). Plus the more popular a link is, the more likely the tweet is to be retweeted, maybe? Does your analysis include retweets?
What does the graph look like if you just count the clicks? If it looks like an exponential function it would indicate that search/#tags/retweets might be having an effect.
Joshua Mathe 12:36 PM on September 27, 2011
But what about Retweets? While 130 characters is okay, using less allows for a higher engagement rate from users who want to share and add their two cents on top of it. Did you take this into consideration> what will ultimately drive more traffic to the link in question?
Jeff 12:40 PM on September 27, 2011
Causation not equal to correlation.
Ryan Smith 1:47 PM on September 27, 2011
I think by using up to 130 characters this gives a little cushion for users retweating. It's hard for me to imagine a large drop off between 130 characters and 140 characters. Tweets are short by design anyway. What's 10 more characters?
Tolson 2:28 PM on September 27, 2011
A little skepticism is always a good thing. I don't doubt the data in this analysis, but I think that it is a bit of a jump to say that people should lengthen their tweets to get more clicks. If you can make your point quickly I don't think that adding unnecessary words will get you better clicks (I think that one of the commenters illustrated this already). What I think might be a more practical way to look at it is: don't be afraid to add a few extra words or characters IF it will help to make your tweet more compelling. Apparently you won't be penalized. But don't go long just for the sake of going long.
Glenn 2:53 PM on September 27, 2011
Pete Austin -- Great point !
Chris Lightner 4:00 PM on September 27, 2011
The steep drop off at 130 makes me think that ReTweets are contributing to these click metrics. If you use the full 140 characters then you don't leave much room for people to retweet without manipulating the tweet itself, and potentially the URL. Leaving those 10 extra characters maximizes the number of characters you use while also allowing for the message to spread through ReTweets, which of course increases the CTR on the links. I'd love to learn more about/discuss the effectiveness of the resulting RTs and how those metrics relate to the followers of the people retweeting, the length of their retweets should the tweet change, etc..
Great post, thanks!
ptaylor98 4:57 PM on September 27, 2011
Interesting study, and the suggestion to remain under 130 characters to allow RT is certainly valid. Obviously though, more info in the tweet, the more RTs.
Adinah 6:39 PM on September 27, 2011
Looking at the chart for, lets say 120 characters, am I believe that the average tweet gets about a 12% click through??? Seems WAY out of line. I doubt if 12% of the followers even see a specific tweet. There must be some other explanation. Perhaps retweets which can indeed be longer when the retweeter adds 'Great post-' or something similar.
Frances 7:41 PM on September 27, 2011
I was also surprised to see this. I guess I shouldn't be emulating Guy Kawasaki's short tweets after all. Good to know!
Josh Blankenship 11:29 PM on September 27, 2011
So, stay between 20-35 or 110 -130 characters! Those appear to be the sweet spots.
Frekklz 1:01 AM on September 28, 2011
I would like to see data on which url shortener gets the most click thrus.
Thanks for this it is good to know! I do believe leaving some space is good for retweeters!
Karen James 4:35 AM on September 28, 2011
Interesting post - I'll have to test it out. How long did it take you to analyse 200,000 tweets!
Maria Peagler 2:03 PM on September 28, 2011
Great analysis, as always Dan. I'm not surprised at findings: I never click on a link unless the person gives some sort of editorial comment or idea of what it's about. My time is too committed.
Carl Natale 9:22 PM on September 28, 2011
I think what creates the clickthroughs is more context not characters. The audience wants to know what they're getting. They're not looking for ambiguity.
Elke Gubbei 11:47 PM on September 28, 2011
Danke für diese Information. Zeigt sich doch, dass etwas mehr an Information gefragt ist.
Grüße von Elke
Thank you for this information. But shows that something more is asked for information. Greetings from Elke
badana 8:44 AM on September 29, 2011
health was very good hands thanks
badana 8:45 AM on September 29, 2011
health was very good hands thanks
AnujShailGupt 9:28 AM on September 29, 2011
Thanks for the intriguing stats.The people moved to twitter because they need less text but those 140 characters means a lot of concised info.
Gregory C. 10:42 PM on September 29, 2011
Definitely shocking, I wonder why the 130 character limit, why does it effect clicks?
Sudhir Kumar 1:57 AM on October 04, 2011
This is really surprising! I was in impression that if you write shorter people tend to click more to see detail. thanks for sharing this info.
Carrie Taaca 9:00 PM on October 06, 2011
Hi Dan,
I have a quick question for you. Can you contact me?
Thanks,
Carrie Taaca
Higher Level Group
John 10:13 AM on October 12, 2011
This totally makes sense! When we started Tripped Media we received a huge increase in the amount of traffic after we accompanied out tweets with a little background and some laughs. Who would've thought?!
Chuck 12:36 PM on October 16, 2011
Because it gives people more information on why they should click it.
Nico Pelletier 3:15 PM on October 26, 2011
I agree with the RT issue. I get way more clicks when my tweets are RT, multiplying my reach.
I keep my tweets around 100 characters, and regularly add PLS RT (like mentionned in the great webinar from HubSpot from a month ago)
Thx