But wait a minute — does this prove that Tuesday or Thursday is the best day to post? Because what this actually shows us is that Tuesday is the most popular day to post (and by a narrow margin at that.) But is it the best? I’m glad you asked…
Answer: It seemingly depends on the company, but there do appear to be a number who publish their most successful posts on Tuesday. Now, whether or not that is because they too believe that Tuesday is the best day to publish and thus save their best posts for then I can’t say. But it won’t matter either way if it isn’t statistically significant.
What about Post-Length?
There doesn’t appear to a very strong relationship between post length and total shares, though you can definitely see a positive trend, especially just after the 1000 word mark.
The median length for all posts is 825 words. For those with >500 shares, it’s 1427 and for >1000 shares, it’s 1609.
So What Did We Learn?
- Tues-Thurs are the most popular days to post but good posts can get a good number of shares on any day.
- Posts should be at least 1000 words but receive diminishing returns after ~1500.
- Twitter and Linkedin are equally the best distribution channels.
- We didn’t get a chance to test this for all posts but we began to see a trend of posts with infographics and videos receiving a large number of shares. This makes sense but was difficult to classify in our current model where we were looking primarily at post length (where videos + infographics have a low amount of words.)
- Looking at channels, timing, and length is a good start but none of that can truly predict how well a post is going to do. We’re going to need to get into the content itself to answer that.
+Commentary: Above is a tremendous post using statistical & predictive modelling to unmask the much heralded claim that Tuesday or Thursday are the best times to post. This is a highly truncated version of just the notable takeaways. If you are a data wonk, then by all means read the whole thing. The point of sharing it here, and it is not cherry-picking, but finding someone, anyone, who dared swim against that inevitable tide, and surely more than a billion articles (all variations in some form of the other to write short ∼350-750 (Buzzfeed/Snackable Content) and not ∼1000-1500 is big news, that should come with fireworks and confetti raining down on everyone.
So much of what is said here cuts across the long held tropes that two days are better than others. Without ever considering, for whom? These types of decisions, and even this predictive modelling is usually gobbled up, like faux science reportage which instill collectively in people these norms, maxims, etc… and yet they can or will be proven false by data all the time. Just not in real-time fact checking.
So we encourage you to first get a handle on when your small sample size, or certainly smaller than the majority of these larger publishers. These things will vary in results, depending largely on your application. If everyone else is posting on Tuesdays won’t your content just be swimming in a tidal swell of stuff? Not the best way to get it to seen. Especially if you are an independent force trying to offer something the rest aren’t.
Again, by rethinking the way we interact with this data and especially focusing on what it tells us about the field as a whole, you have to chart your own course. Don’t let these ideas become a part of your bias, simply focus on doing the best you can, connecting as many dots in your network as possible, and turn out good content.
The rest of this will matter very little if it is posted on Tuesday or Saturday. This sampling didn’t consider content either as it clearly states. So removing these myths can sometimes encourage people to do what they are doing with more attention to the output, and less on the timing. Certainly timing & strategy play into it, but not in a cookie-cutter fashion of posting on agreed upon times/days. So go forth liberated to do what you are already doing with the confidence that the real thing that will matter will be content worth sharing, not the usual timing of large outlets.