DISQUS

evbart.com: Facebook vs. Twitter: Round two with URL shorteners as the judge

  • @hvhuck on twitter · 6 months ago
    Fascinating, elegant study. It also mirrors my interaction with both media and thus rings true for me.
  • evbart · 6 months ago
    Thanks. Can't claim that its ground breaking research :-) but I was a bit
    surprised to see how much Twitter was out performing Facebook.
    Not a good sign for brands really banking on user engagement through
    facebook!
  • Nathan T. Baker (NateNe.ws) · 6 months ago
    Get you some of that twitter. Since Twitter is still in a Wild West stage, it will be interesting how this tappers off as twitter becomes more mainstream.

    I think the bare bones idea of twitter focuses people on interacting and sharing. I'll get lost looking at pictures for hours in facebook, whereas in twitter, I'm communicating by default.
  • Antonella · 6 months ago
    Interesting but I would also like to see what happened after the click.
    Twitter users are more likely to click on links but do they "convert"?
  • evbart · 6 months ago
    @ Nathan I agree twitter is more focused on interacting/engaging, while
    facebook is more focused on your friends.
    @ Antonella in this case they were just going to a page that had a directory
    of apps built on the Kiva API, so we weren't trying to drive a particular
    activity. Would be great to see how users from Twitter convert into users
    on Kiva, but I don't have those stats :-)
  • sue_anne · 6 months ago
    I agree with Antonella. I like these stats for what you were obviously trying to measure, but it would be interesting to see if there was a particular action which traffic source was more conducive to that action.
  • aweissman · 6 months ago
    Great study - it's just one test but interesting to see nonetheless. Thx for publishing
  • AlexSchleber · 6 months ago
    The result is not that surprising, since Twitter users have been "trained" to practice openness, engagement, and most importantly propagation via Retweets (which Facebook "Likes" cannot match as they don't create the same degree of surfacing, largely because FB users don't (yet) expect to interact with other users' Friends of Friends or their "fanned" FB Pages, etc. to the same degree as on Twitter's wide-open platform).

    Granted Zuckerberg is hard at work at retraining them...

    Nonetheless, smart bit of research using a relatively simple tool like bit.ly. Shows how important it is to get actual confirmation/feedback via stats on how postings to the various communities are received.

    I have been arguing this over on FriendFeed for a number of weeks now, as there currently is no good way to measure clicks on either links shown internally on FF, or the shortened ff.im links shown if FF actions/postings are forwarded to Twitter.

    Ever since I've been using FriendFeed in an accelerated fashion starting about two months ago, this has been an issue, since I have no real way of knowing to what degree my Twitter followers are engaging with my FF materials. Anecdotally, it seems that click-throughs on ff.im are currently much lower than on bit.ly links, in part simply because fewer users are familiar with them, asf.

    Here is the thread:
    http://friendfeed.com/friendfeed-feedback/481ec...
  • euwyn · 4 months ago
    Evan,

    This is excellent! I'll try to run a similar test over at Wokai. I'm actually not that surprised by the results. Even from an interface standpoint, links are much more prominent on twitter.com and any other Twitter app. Links are a little lost on FB, amidst all the pictures, "Joe likes this", the blue comment boxes, long story snippets, etc. I would also guess a large chunk of folks like me still use FB like the good old days, mainly for messaging and chatting with friends and not necessary for its News Feed, no matter how much FB tries to push that on us.

    - Euwyn
  • nabeel · 3 months ago
    Did you take account of bots?

    We ran similar studies over at Loudcrowd, and at first were very pleasantly surprised by Twitters performance.

    However, we became skeptical when we found the bounce rate incredibly high, and it turned out a big portion of the traffic was bots, enough that FB became the better channel. I should do a post on it.
  • evbart · 3 months ago
    No, I didn't. How would you track which part of the traffic comes from
    bots?
    Why are bots clicking through your tweets?
  • nabeel · 3 months ago
    The base problem is that FB is private, and Twitter is public -- bots troll the Internet, it's just what they do.

    Unfortunately, there isn't an easy way to clearly find out if it is a bot, some report themselves, some don't execute javascript (so google analytics won't track them), but some do and you need to look at capabilities of the user agent (does the browser have Flash? is it running at 640x480 resolution?)

    We saw as much as 75% of twitter inbound traffic fell into one of those three buckets and ended up being bots. This is general noise on a major website, but with an individualized bit.ly URL that gets less than 500 clicks a day it can easily be a big %.
  • evbart · 3 months ago
    You should do a post on this, it would be interesting to see what percentage
    of traffic this makes up. Definitely put the link up here if you do.
  • evbart · 3 months ago
    You should do a post on this, it would be interesting to see what percentage
    of traffic this makes up. Definitely put the link up here if you do.