When publishing Twitter statistics, one of the most common questions I'm asked is about the effect of setting a profile picture in your Twitter account.
I dug into Twitter Grader data and analyzed nearly 9 million Twitter profiles to produce the graph below.
It doesn't take much, but if you want to get followers on Twitter, it's a good idea to upload a picture of yourself.
David Belden 2:08 AM on April 06, 2010
Does a company logo count?
Victor Canada 2:26 AM on April 06, 2010
Part of the equation is that when there is a picture the person is deciding if they are interested in what you have to say. If there is a logo, the person is deciding if the company has value for them. The tweets could be the same but the process for deciding is different. No picture at all is anti-social. Unusual pictures may put someone off or attract exactly who you wanted.
Jana Sheeder 2:27 AM on April 06, 2010
Great question, David. I wonder, also. I don't know if a corporate logo is considered the same as a photo of a person! I wonder if Twitter Grader's statistics delve deeply enough to determine that difference? Thanks, Jana
Iain Duigan 2:35 AM on April 06, 2010
Dan, you are so right. In 36 hours after I put up a personal photograph of myself I went from 61 followers to 195. Although I'm a little disillusioned as I thought it was my handsome face that did it!
Dara Bell 5:59 AM on April 06, 2010
Seems natural, I always book a development event from looking at the people running it. You can check backgrounds and other things but often my first impression is correct.
I do go thru background checks on bigger budget things like 10,000 grand courses or services like graphc design but again often the work itself gives more cluses to me, not reams of marketed or promotional content.
Dara Bell 6:00 AM on April 06, 2010
Seems natural, I always book a development event from looking at the people running it. You can check backgrounds and other things but often my first impression is correct.
I do go thru background checks on bigger budget things like 10,000 grand courses or services like graphc design but again often the work itself gives more cluses to me, not reams of marketed or promotional content.
Dara Bell 6:00 AM on April 06, 2010
Seems natural, I always book a development event from looking at the people running it. You can check backgrounds and other things but often my first impression is correct.
I do go thru background checks on bigger budget things like 10,000 grand courses or services like graphc design but again often the work itself gives more cluses to me, not reams of marketed or promotional content.
John McTigue 7:46 AM on April 06, 2010
There are so many spammers and scammers, it makes sense that a real person would appear to be more trustworthy and gain more followers. I think company logos are ok if the tweets are clearly from the company, not a single person. In addition to avoiding strange pictures or twitter logos, I also avoid strange names that seem to have nothing to do with a person or company.
Michelle Quillin 9:51 AM on April 06, 2010
This is something we've really wrestled with -- do we stick with the modified company logo or use a personal picture of me, since I handle all the social media for New England Multimedia? We did customize our Twitter background, including the real company logo and a photo of Scott & I back-to-back in a superhero pose (haha!), but I have this nagging feeling that our Twitter profile pic is just too impersonal.
The thing is, the modified company logo stands out in a Twitter stream full of human faces.
Maybe I'll try an experiment and see what happens.
Michelle Quillin for New England Multimedia and Q Web Consulting
http://twitter.com/NEMultimedia
Jonathon 9:55 AM on April 06, 2010
Yes, could you please comment about logos vs photos? I'm wondering if I should change the http://www.twitter.com/photoglow profile pic to a photo of myself or leave it as the companey logo. Cheers.
Dara Bell 10:54 AM on April 06, 2010
Myself I am writing a book on altruism. The superhero sort of represents this but does it say the wrong thing.
Would a headshot be better. Your opinions would be appreciated as I am going through a personal brnading makeover.
Patti Moran 12:08 PM on April 06, 2010
Interesting...and very timely for me. I had been using my company logo...I am self-employed, so it's just me. I have a very distinctive logo which, as a graphic designer, I'm proud of and which also serves as a nice example of my abilities. Also, as Michelle mentioned. the logo really stands out in the sea of face shots in a Twitter stream...that can't be a bad thing, can it?
Even so, I'd been pondering the logo vs. face thing, so just yesterday I polled my Twitter followers on it. Only a handful replied, but they all said they'd prefer a face over a logo. So I changed it to my face. Now I can hardly find my own Tweets in a stream!
I'm still not convinced that a face is better than a logo, so I'd LOVE some evidence, like your face vs. no face chart.
Jeff 12:09 PM on April 06, 2010
I have the same question as others. Would it be more effective for a company to use a logo or the photo of the person actually behind the logo who's interacting on behalf of the organization? Or a combination of the two?
Simon 12:12 PM on April 06, 2010
What have you done to control for variables? Correlation does not mean causation. Twitter accounts with default avatars also seem to often be spammers and/or newbies who don't tweet much or effectively.
Patti Moran 12:19 PM on April 06, 2010
LOL, poor Dan. He wrote a really interesting article, and no one is commenting on what's IN the article, only what's not in it. Dan...it WAS a really interesting post. The difference between followers in profiles with or without photos is really remarkable.
Whether or not those with photos are using logos, or whether they're spammers (most of the spammer's I've seen have photos), it's clear that using any kind of photo is far better than a twitter default avatar.
Dan Zarrella 1:04 PM on April 06, 2010
We're working on ways to do a followup post about the content of the images: ie if having a picture of a real person has any effect on follower numbers.
Michelle Quillin 4:56 PM on April 06, 2010
Patti, good point!! Thanks, Dan!
About the superhero pose -- we didn't intend for it to look like that; it just turned out that way. It actually kinda looks like we could be the stars of one of those cop shows. We're the good guys, at any rate. "At your service!" ;o)
Thinking of incorporating our logo into the background of an actual picture of me, or of both of us. Maybe that'll help with the "stand out" issue.
Dan, looking forward to updates with new info.
Michelle Quillin for New England Multimedia and Q Web Consulting
http://twitter.com/NEMultimedia
Mark 6:51 PM on April 06, 2010
This can't be serious? That graph does not prove 'The effect of profile pictures on followers'. All it shows is that the people who use twitter regularly and therefore develop a following are also the people who bother to put up a picture. The people who sign up to twitter and get bored and have few followers are likely not to make a picture. The follower numbers aren't directly to do with the picture, though of course a pic helps.
Alphabetix 10:33 PM on April 06, 2010
I think Mark makes a good point that active tweeters (and those more likely to have followers) are often those who take the time to use a profile picture. Correlation not causation, eh?
But, also I do wonder about the logo thing. It seems to me that if you're tweeting for your business, it only makes sense to use a logo, especially if there are a number of people who tweet on that account.
Victor Canada 11:01 PM on April 06, 2010
christian louboutin spamming a blog about Twitter account best practices. Seriously? So sad. Please find someplace else to sell your shoes. Ridiculous!
Peter Low 6:03 AM on April 08, 2010
Yes I would ask myself who am I making friends with if there is no picture of the person that is asking me to become friends with.
Ian Greenleigh 12:11 PM on April 08, 2010
Dan-
HubSpot's blog is one of THE models, as far as I'm concerned.
My one point of critique so far is that I'd like to see less presumptions regarding causality, or if you're going to say things like "effect of", providing more regarding your methodology and data used.
The correlation alone is interesting enough without attributing causation.