24 Apr 2013
Best time to post : why it’s a lot harder to predict than it seems + Nitro announcement
We’ve received a lot of requests to add “Best day/hour to post” in our Facebook analytics solution.
This could have been easy to integrate if we had decided to follow our predecessors’ paths : take your previous posts, do some basic math like average/median score and there you have the best hours and best days to post.
Problem
Results are not accurate if you do it this way. And yet it seems to be a pretty intuitive approach to do the prediction.
Why is it not statistically accurate ?
Here is a quick example.
Based on this approach, the following post should have tremendous success if posted at the right time and hour.
Hey! Isn’t it suppose to work like a charm ? I’ve posted it at the right time and hour!
As you can see, the problem with this approach is that it only takes into account the hour and day as factors of success.
Let’s say that the tool recommends you to post on Friday at 4pm.
What if :
- You mainly post during Friday afternoons (without you even noticing it) or
- On Friday afternoons, you mainly post photos while the rest of the week, you generally post status updates or videos, which in your case, don’t generate as much engagement as photos or
- On Friday, your posts are about fun activities during week-end and they resonate well with your audience
Enough said. You can’t resume the success of a post to timing only as this success is based on a lot of different factors.
We’ve identified more than 60… and hour and day are only two of them!
There are factors related to the post itself (type, topic, length of message…) or factors related to the context (success of previous posts, competition in the news feed, publication frequency…).
So what’s the solution?
Posting the same message every day, every half-hour for two non-consecutive months in order to isolate the factors other than hour and day that affected your results.
We know it’s quite impossible (but if you try, we’d like to know the results!) so we’ve devised a unique algorithm that identify the posts which have been affected by the timing more than the other factors.
Sneak Peek – Introducing Wisemetrics Nitro
This algorithm is a core part of our brand new predictive posting technology some of our clients are currently beta-testing.
This technology, named Nitro will automatically optimize your posting on Facebook.
Stay tuned!
