I admit, I was intensely skeptical that How to Schedule Your Twitter Account Posts and Twitter Auto-Scheduler Reloaded would really bring my site much traffic at all. My objective was Automating My Life to achieve low maintenance blogging for myself and of course high ROI in terms of traffic. There was no way this programming blogger was giving a dime to a company like CrowdFire. It turns out some “online strategy” blogs you read give you amazing, solid advice. Vicky’s approach for #5 gave me the idea for How to Schedule Your Twitter Account Posts and my site’s traffic exploded. Thanks Vicky!
Google Analytics and AdSense
When I say exploded I mean trending upward day after day. As you can tell my site has done great after executing that little Python script. I’m not really one to boast much but this graph is epic. What you see in the graph is Feb to May of traffic data. I had a few false starts there at the beginning but after working through some deployment issues I was off to the races. Another cool thing is that ~95% of my traffic is always new, but my issue with that is I only have new traffic. It would be nice to have returning customers. The average “reading time” for the 4 months is 3:33 minutes (88% bounce rate). That means for each new session a user has spent 3 minutes of their day reading 1.24 pages per visit. I think that’s outstanding! Adsense is still struggling to make boatloads of money for me ($0.17 in 3 months) but I did get 2 ad clicks in April. Adsense has a threshold of $100 before it pays out but an upward trend is okay with me. That’ll do donkey, that’ll do.
Another metric that I have found most useful is the PageRank of your website. Google Analytics tracks user behavior and ranks your pages by most visits. Stream Spotify From Python is still the Top Winner. Also, it seems that an AWS Datacenter in Ireland really loves that post. I’ve also been visited by Labs.Spotify domain as well
which obviously means I’m legit now.
Another cool new feature is called Behavior Flow. Behavior Flow is Google’s attempt to identify what a user does when they access your website. I will translate for you.
You can see from the blocks that 125 users landed on my Stream Spotify From Python post then 9 of them decided “Hey, I’ll take a look at the main web-page” as denoted by “/” in the block. This choice to click to another webpage is the 1st interaction. After discovering another post that may interest them, 1 person clicked on to How to Schedule Your Twitter Account Posts. Amazing!
My Twitter account has really done great since Feb.
since I basically started at zero. In the past 30 days I have gotten 40 retweets, 62 likes, and 16 link clicks. My little Data Scientist heart is just fluttering away. All of this with one Python script Twitter Auto-Scheduler Reloaded. I now have 48 followers although some I know for sure are Twitterbots and SEO companies. However, that isn’t such a bad thing is it? I’m on the fence.
Optimize My Twitter Poster?
Since you know I’m a Data Geek and an automation lover, my next post to sunset the Twitter blog post series is to create a “Data-driven Twitter Poster”. That’s right, my plan is to use Google’s Analytics and automatically decide when the most optimal days, times and frequencies for posting are. I can also predict in the future how many clicks I’ll get on a given day using some sort of regression. Maybe I get the most ROI on Mother’s Day (you wouldn’t believe the Twitter interactions on that day!!) or maybe the best time frame for my Spotify post is Tuesday at 5:03 pm. With data you never know.