Are you guys cool if I keep Twitter spamming you? #tweetoldpost

Is Twitter the new spam?

I’ve had this problem figuring out how to use Twitter as an awareness tool for my blog. Outside of Google search and people directly typing in the URL (including RSS or bookmarking it), Twitter is by far the next greatest source of traffic. I personally use Twitter more as a communication forum than I do anything else, but the fact that it was driving a lot of traffic to my blog was not lost on me.

The problem I’ve always had is that Twitter is by and large an immediate medium. If you happen to tweet when someone is using Twitter, you’re golden. If you tweet a few hours before they get on Twitter, there’s a good chance they’ll never see your message. And what about your friends not in your time zone? I’ve found that if I tweet my blog posts around 8am EST, it gets the most visibility, but I’m certain my friends on the West coast aren’t up at 5am on Twitter and likely are missing the tweet.

The question then became, do I tweet a blog post multiple times during a day so I get the most exposure while risking those that had already seen it being annoyed? I was always hesitant to do that.

Then I was introduced to the WordPress plugin Tweet Old Post. Here is the description:

Tweet Old Posts randomly picks your older post based on the interval specified by you. The primary function of this plugin is to promote older blog posts by tweeting about them and getting more traffic.

A few people I knew had used it and loved it so I figured I’d give it a shot. I was careful to adjust the settings so that I wasn’t blasting old blog posts out constantly. I tried 8 hour increments and set it to only tweet posts from the last three months. And then I turned it on.

Over a month and a half period, compared to the previous month and a half, some very interesting things happened:

  • Overall traffic to my blog increased by 10%. I actually thought it was more until I ran the report. What I found the most exciting was that weekend traffic rose considerably since I typically don’t blog on the weekends. What’s interesting to note is during that time I actually blogged significantly less (was traveling and very busy at work). So traffic going up by 10% when I was blogging less was pretty big.
  • The percentage of traffic I received from Google rose by 60%. This surprised me and only further reinforces the influence that Twitter has against search results.
  • Average time on site dropped by 17% and average pages per visit went down by 7.6%. I’m not surprised by this as I had considerably more first-time visitors to the blog than recurring during this period.
  • My traffic smoothed out considerably. Instead of having spikes on days I blogged and large dips the days I didn’t, the graph became much more consistent. See the image below, the green line is the previous traffic and the blue line is the traffic with Tweet Old Post.

Blog traffic. Green line is before Tweet Old Post, the blue is with Tweet Old Post.

Again, I think when looking at these results you have to consider that I went from blogging 3-4 times a week to 1-2 times a week during this period. Not on purpose, but I happened to be busier with work plus I honestly think Tweet Old Post made me lazier as I saw traffic coming to the blog regardless of whether or not I was blogging. So the results are more substantial than what is depicted above.

I was pretty happy with the results of Tweet Old Post but I was paranoid about whether or not my Twitter friends were feeling like I was spamming them. Then last week a few of my more honest Twitter friends raised their hands (thanks @rkischuk for starting that conversation) and validated my concern, that they didn’t like the new strategy of tweeting old stuff out.

So the dilemma at this point is, do I stop tweeting old stuff out for fear of turning off my more loyal readers/twitter friends, or do I continue tweeting out older posts so that more people see the content I’ve written and I grow my follower base…only to them possibly annoy them with the old posts (#ironic).

At this point I’ve turned off Tweet Old Post until I figure out a solution. I’m very torn between turning it back on vs. not. I could tag every old tweet with #fromthearchives or something like that. I’ve also considered every few months turning it on for a 2-3 week period to seed recent content that I think might have been missed, and then turning it off again.

Looking for advice…

7 Comments

  1. Drew Hawkins on September 7, 2011 at 1:09 pm

    I know I’ve said it before but I think it’s like any automated social service. Depending on how you use it depends on effectiveness. The decreased frequency helps but also using posts that are no more than X days old or categorizing a “best of” section of posts. The second idea would be labor intensive but could separate more “timeless” posts than time-sensitive ones. 



  2. Sherry Heyl on September 8, 2011 at 7:53 pm

    This is a good example that increased traffic is not always the best measurement. It really depends on if people who came to the site was satisfied with what they saw. And in the bigger picture I would not change my strategy based on one person’s honest opinion. One person may read every post of yours and is excited to see when you tweet your latest post. Others may be interested in content you wrote several months ago but did not see it until it was tweeted. This strategy might have also led to new readers finding you, as you point out…I do not think I would automate this but perhaps hand pick what old posts I would want to schedule to be tweeted out. 



  3. Jim Gordon on September 9, 2011 at 2:47 am

    For one, you’re probably seeing your Pages Per Visit go down because the bounce rate will inherently increase when you’re linking people to older content that doesn’t necessarily qualify a lead like fresh content.  Google Analytics doesn’t factor in bounces when calculating Time on Site, so you’re probably just seeing people recognizing it is old content, clicking to the Home Page, and leaving which is driving that stat down.  You may also have people being driven off of the site by links within your blog post.  If you had click tags, your bounce rate would take a serious nosedive.

    With regards to post frequency – why not post the same article twice in a day?  Those who diligently follow you will know when a new post is up and also recognize when an older article is posted.  The issue you’re running into is that you have 2 groups of people – morning people and afternoon people (and a mix of people in this Venn diagram).  You want to maximize the exposure of your new content with your audience without being redundant.  By posting old articles, you’re potentially experiencing a higher bounce rate and fewer PPV.  Post the same thing later in the day and see what happens.  The people who read it in the morning won’t click on it and bounce because they have already seen it and those who haven’t will click on it and stick.

    Problem solved?



  4. Adam Steinberg on September 9, 2011 at 1:20 pm

    What’s the loyalty score of new readers that find your blog via the auto-tweets?  If the increase in new-reader engagements offsets the loss from existing readers, I think you have your answer.  This is assuming that your most important metrics is readers/page views.  

    Personally, I like the auto-tweeting.  Helps me discover previous blog posts I may have missed.  



  5. Jeff Hilimire on September 9, 2011 at 2:47 pm

    Thanks Sherry. I agree, the ultimate would be to hand pick out blogs and schedule them. I think that might be where I end up with this…



  6. Jeff Hilimire on September 9, 2011 at 3:39 pm

    Tweeting out a blog post a second time in a day is definitely an option. Here’s an article I just saw on it: https://selnd.com/q9JH7E



  7. Where I landed with tweeting old blog posts on September 12, 2011 at 4:48 am

    […] Last week I posed the question of whether or not I should be tweeting out older blog posts. You can see the post and the discussion here. […]



Leave a Comment