The Most Influential Celebrities on Twitter

Who do you think has more influence, Kim Kardashian or Martha Stewart? If you picked the jail-bird, you'd be wrong: homemaking expertise has nothing on bold badonkadonk. At least on Twitter.

According to a new analysis by social media guru Dan Zarrella (who previously helped us learn Steve Jobs' presentation techniques), the Kardashian sisters are some of the most influential celebrities on Twitter. By studying click-through rates—the amount of clicks a tweeted link receives—Zarrella was able to graph celebs' influence on their followers (see chart below). Some of the names are not so surprising (Ashton Kutcher), but others, including British actor Stephen Fry and poet laureate "Weird Al" Yankovic, may come as more of a shock.

Influence comes in many forms. The Kardashian sisters may gain more clicks because of the many provocative photographs they tweet. On the other hand, someone like Lance Armstrong may reach more fans because of his powerful message. And then there are those who have a perhaps more "nerdy" following, such as "Weird Al," Stephen Fry, and Felicia Day. "Felicia is probably high because she's a web star so her followers are more web savvy and likely to follow her links," Zarrella says.

"Alyssa Milano is my favorite example," he continues. "She has less followers than most of the others on the list, but drives huge traffic."

Zarrella's findings certainly echo a recent study by HP Labs, which concluded that influence online had nothing to do with sheer numbers—especially on Twitter—but how much users could be swayed from the passive to the active.

Don't tell that to Biz Stone.

Add New Comment


  • Jennifer

    Oh, I think that you need to re-define what's influence word mean, because I don't think gaining more clicks is the same with influence :(

  • Richard

    I think you really need to look for a better expert for getting data for your pieces. I originally commented on the sources data collection methods and how suspect they actually are. He's giving complete credit for all clicks to people no matter if they are the originator or not. The 2 largest of Khloe's links are ones that were RT'd about her sisters baby pics in a magazine and her other sister's wax figure. Here are the stats -

    obviously you can't tell how many of those were Khloe's, she was RTing the main link others had been using. And those 2 links count for 400k of the clicks he had attributed to her, no small number.

    Now when I tried to explain how this stuff works to him and how things like the Twitter API worked, he just deleted my comments and tweets to me. You would think he would be appreciative to see how his numbers really don't do a good job of assessing the situation and that if he had a full understanding he would have seen that to begin with before putting it together.

    Please in the future use experts in the field instead of going with people who just throw what they think is good research out there. Having data like this is pointless and no positive purpose for those who are trying to learn about what's actually going on.

  • Olivier Redmont

    I'm not sure you understand the concept of influence. Whether or not she was RTing the main link that others had been using is irrelevant. The fact is that they arrived to the photos or whatever the link may be as a direct result of following or landing on Khloe's twitter. On purely a percentage basis, the data makes COMPLETE sense. It's just a matter of the reader being able to understand the data properly. In fact, he clearly offers plausible reasons as to why these people are on the list (cult followings, etc.). I disagree with you and think that there is positive purpose in the data if understood properly.

  • Joe Porter

    First off, Fail for making me register to put a comment. I would normally just walk away but need to make sure i understand this data.

    A one word Correct or Incorrect answer from Olivier Redmont will do just fine for each question.

    1.) The data is gathered on shortened links that a celebrity tweets and it doesnt matter if they created it or not. Correct or Incorrect?

    2.) The counts are based upon the TOTAL lifespan of the the shortened link... whether it was clicked on a celebrity tweet or anybodies for that matter. Correct or Incorrect?

    to be continued....

  • Richard

    I'm pretty sure I have a better understanding of what these numbers mean and what the concept of influence is.

    Here is an easy to use examples because of the time difference between the original tweet and the RT so you can understand there are significant issues with these numbers.

    RT -
    Stats -

    The orig tweet came at 1:31PM Sept 3rd and the RT happened between 1:48 and 2:42 AM on Sept 4th. Looking at the stats you can see that 13.7k of the 24k clicks (57%) happened on the 3rd. But through this method you are giving the person 100% credit for it because you don't know. Unfortunately we can't get more precise info but to give someone 100% credit because you can't tell instead of throwing the data out because it's useless means nothing. Overall this data means absolutely nothing. And saying because it now in percents it means nothing when none of the data can be tied directly to the person you attribute it too and it can be any portion of that traffic all the way down to 0%.

    There is nothing to be gained from the use of this data and I don't see how you can say there is anything positive to be taken from it other then RT links with lots of clicks so you look like you are influential to the unaware who write "studies" about it.

    As well your "plausible reasons" of a cult following isn't backed up by this for why it's so high because you don't actually know that any of these people clicked through them. And this is especially true in the case of Khloe as she has two 4+ mill follower sisters that she's RTing about that she then is getting credit for as if she created.

    Right now my % is 5000+ because I had tweeted the previous links as examples. Does that mean anything about my influence? No. As does any of these random click throughs. What I'm completely lost on is how obvious it is that you can't make these kind of statements based on numbers like these yet you are defending it like it can be because it's %'s.

    I'm not saying that there are not cases where they do create content or are the creator of the link that you can then see their influence from, but you have to manually pull those out over just taking blanket numbers to get a better view. You can't just take them all and expect that to be good enough because you "average it" when a lot of celebrity sharing is really rehashed from others so definitely tainted.

  • Dave Woodson

    I see how they think they have influence, but who arr they influencing. Certainly, not me, the 40+, white guy which has been the bread and butter of any marketing firm. You should be looking at the people down the line, not the main celebrity. We all know at this point she goes to the highest bidder. So, therefore, she is often ignored more.

  • Isobel Kramen

    Why do we call them celebrities? Many of the names on the list were not recognizable to me. My definition of a "celebrity" is someone who has demonstrated outstanding talent or knowledge to be known by all by photo or name. Where is Dame Liz, Dr. Bill Cosby, Fidel Castro, Dolly Parton, etc etc and so forth ?

  • Nate O'Shaughnessey

    "celebrity" in today's culture applies to anyone who many people know. I also do not even know half the names on the list. But appearantly many people do. To be honest with you, I don't know who "Dame Liz" is, and does Fidel Castro "demonstrate outstanding talent or knowledge"? Obviously, you beleive so, and that is fine. But "celebrity" is a term that is both subjective and objective. the sheer numbers show that these individuals are known by many, but many of them are not known by you, or by me, so we wouldn't have thought of them as celebrities, that's the subjective part. It's like the terms "respectable" or "intelligent" it applies to those whom you view as that...but that is often different from other's opinions of the same people with those terms.

  • vkodass

    I agree with you but from my personal experience, it bought 10000 Twitter followers from and they added them to my page in a little over 2 month. All the followers appeared to be real and some of them turned out to be great customers.

  • vkodass

    I agree with you but from my personal experience, it bought 10000 Twitter followers from and they added them to my page in a little over 2 month. All the followers appeared to be real and some of them turned out to be great customers.