I was digging through this blog’s stats this morning in Google Analytics, and I noticed something interesting. As I mentioned here a couple of weeks ago, one of my goals for 2013 is to blog more often. A big reason why is because I want to increase the traffic here, and so far in January traffic is up a bit, and that’s to be expected from a more frequent posting pattern.
So what I want to do as the traffic increases is understand which traffic sources are driving that increase. One of them is Twitter, which is by far the social site that sends me the most referral traffic. I started looking at the individual posts in 2013 that have received the most traffic from Twitter, and I noticed something: They received almost all of their referral traffic from Twitter in the first 2-3 days, then almost nothing.
Here’s the number of Twitter referrals for the most popular post so far this month, ‘Done’ is Better Than ‘Perfect’ When it Comes to Blogging, which was published on Jan 1st:
As you can see, good for 3 days, then traffic volume falls off a cliff.
Here’s the traffic pattern for the 3rd most popular post (#2 was one that I linked to for multiple days as a #blogchat topic and it skewed the results a bit I think), which was The Biggest Mistake Companies Make When Engaging Their Fans Via Social Media published on Jan 8th:
Same pattern, traffic for 3 days, then it dies. There was a bump back on the 14th, and that was the same day the post was listed as a link on a New York Times article and a few people RTed the post on Twitter, I am assuming after they found it via the NYT link.
Finally, here’s the 4th most popular post, Should You Remove the Dates From Your Blog Posts?, on Jan 14th:
Two days, then flatline.
There’s a few takeaways I have from this:
1 – These stats suggest that when we are on Twitter, links have a very short window to get our attention. I don’t think that’s a huge revelation.
2 – If we want to leverage Twitter as a platform to draw attention to our writings, then it pays to focus on fewer topics versus more. If we are only paying attention in short bursts on Twitter, then it helps if you are consistently giving us the same or similar topics to look at.
3 – Since our attention spans are so short, it means we can cover the same topics repeatedly. This is where I think you can really leverage Twitter as a channel to drive big referral traffic back to your blog. By blogging frequently, and by frequently blogging about the same or similar topics. Notice from the above graphs that the 3 posts were published on the 1st, the 8th and the 14th. Pretty spread out, but what if those same posts were published three days in a row? The spike in traffic from each would overlap and by the 2nd and 3rd days, referral traffic as a whole from Twitter would be pretty high.
4 – If you plan on having Twitter be a driver of traffic to your blog, you probably need to publish a new post at least every 2-3 days. If we assume that the average blog post has a lifespan of 2-3 days on Twitter, publishing a new blog post every month isn’t going to do much for your referral traffic from the site.
So that’s something to think about. I think for me what I would like to see happen is to find a way to not only extend that average lifespan to say 3-4 days, but to also chain together posts that bring in higher amounts of referral traffic from Twitter on a more regular basis.
I would also be interested in hearing what the rest of you are seeing with the referral traffic from Twitter to your posts. Are you seeing most of your traffic coming in the first 2-3 days as well? Do you see a longer range? Shorter?
Hi Mack,
I’ve observed the same Twitter spike effect, and one take-away I’d add is that I’ve learned to optimize that narrow window a little more wisely by having conversion and other click-test ducks in a row before I share the post. If there’s a specific action you’re hoping readers will take when they arrive at your page, make sure you’ve got your links in place, that the links are correct, and *maybe* you’ve even taken the time to test them!
Hi Penina! Relevant calls-to-action are something I always struggle with doing properly. I’m doing a bit better with it concerning my book, and my posts here have actually driven a few sales, although I can’t tell how many since they are pre-orders, I will get a better idea once the book is out in April. But great reminder that there’s so many factors in what works and what doesn’t!
Very interesting. Although for me, the reverse is true. When I dont post, I get more traffic. I have tried to look at the differences in types of traffic and even the source, but it’s all very consistent with exception of volume. Trying to think of a test I could do and would love your thoughts. Thanks Mack
So when you don’t post blog posts, or when you don’t post links to them on Twitter? As far as getting more traffic?
Assuming you mean you get more traffic to posts when you don’t post them to Twitter, are you linking out to other blogger’s posts at the same time? I have found in the past that if I link out to new posts from other bloggers without linking to mine, that it will still drive traffic from Twitter because some of those bloggers that I link to will go to my blog and find my new post and give me a reciprocal tweet.
To be fair, I’ve found I get MORE traffic if I follow the same pattern and link to my own blog post as well. For example, link to 10 straight blog posts by different writers, and maybe the 7th link would be to my new post. By the time I get to the end of my list, the first bloggers I linked to will likely either be thanking me and/or linking to my blog post in return.
What is the the traffic pattern when you resend the same post about five days later? Do you even recommend tweeting out your same article two or even three times? I believe Jeff Bullas uses this approach.
Dan what I first do is link out to the post on the day I publish it. And I’ve sent out tweets linking to new posts as many as 5 times in a day, as long as you spread them out by at least a couple of hours, no one really balks.
Then I will go back after a few days and if a post was REALLY popular, I will link to it again after a few days. Another thing I should be doing more of is linking to my ‘all-time best’ posts at least once a week. Pick say the Top 10 posts you’ve ever written, the ones that got a ton of comments and everyone enjoyed them, and link to 1-2 of them a week.
Hi Mack, that is a very interesting approach to analyze traffic source. I own several websites and blogs but I rarely go into details and track specifically to find out how long the stream from twitter actually last after I’ve submitted it. Nevertheless just a quick look-through on my analytics account tells me that most of them actually last about the same number of days as yours. Some as short as one day but on some occasions, I don’t see traffic coming in at all. I don’t know how to explain further but your explanation seems to be valid. On another note, do you think that you ought to include the idea that it might also be influenced by the number of new twitter followers that you’ve just received? Any method to track if this is one of the contributing factors?
Hi May! The new Twitter followers is an interesting idea, but what I find is that most of my new Twitter followers are new to Twitter as well. So they have very small followings, and if they RT one of my posts it’s going to their 200 followers, and probably few of them know who I am, so it typically doesn’t drive a lot of traffic.
What DOES drive traffic (as you can guess) is when someone influential RTs one of my posts. For example I love when Jay Baer RTs my posts because it’s like ‘social proof’ for everyone else that the content is valuable, and I typically get a ‘chain reaction’ of RTs after Jay RTs me. I’ve seen the same thing from Shannon Paul and I did from Beth Harte when she was active on Twitter.
Certain people have high levels of influence over their followers. And what I’ve found is that the people that seem to be very influential on Twitter as far as sending traffic to my blog, are often people with lower Klout scores. Which is interesting.
I should so not be stalking your post comments right now (we’ve got to head out!), but couldn’t resist repeating this:
“the people that seem to be very influential on Twitter as far as sending traffic to my blog, are often people with lower Klout scores.”
Sparks for a new post?