At what time of day is the best moment to promote the website?

Question

For a while now I've been maintaining a CogSci.SE facebook page (facebook-link). At the moment, every Monday at 20:00 CEST (UTC+2), i.e 18:00 GMT and 13:00 PDT (UTC-7), I post a question on the page. Unfortunately, after two months of work, the reach of these posts is not as good as I hoped for. Without further liking and sharing, we reach about 50 people. I believe this may be because of the time that I post the questions.

I chose for the current moment because I knew for sure that everybody would be awake at this moment. I believed that western Europe and the entire US would most likely be our best bet. However, this makes it difficult to time the post just right, because I don't know when people check their Facebook most often.

What would be a better moment for posting these questions?

In order to answer this questions we must determine on which area we will focus our attention. Where can we find the most audience, and at what time in day (morning, afternoon, evening) do we need to post the question?

Answer 1

I don't know much about Facebook, but in the past I've used Reddit to promote both my own blog and StackExchanges like CogSci and cstheory. I find that Reddit can drive a lot of traffic to the site, but the comment threads often splinter with a lot of good feedback coming on Reddit itself rather than being posted as answers or comments here. So I am not sure if this sort of outreach was useful for growing the active user-base.

I've found that on Reddit the highest attention is drawn on weekdays between 8 and 9 in the morning, Eastern US time. It seems that a lot of people are checking their phones on their commutes to work or right after arriving at work? Also, if the posts get significant attention in the US East morning rush, they often see continued attention from Europe/UK after-work commute and getting home time (roughly 2 to 4 hours later).

I am not sure which weekday is the best, and I don't understand the timing on weekends. I also don't know if any of this translates to Facebook.

Answer 2

We are now testing different days and different hours to see whether the amount of views and actions (clicks, shares, likes etc) increases. We aim to test each period for two months. Below is the list of tested moments and their descriptives:

1. Monday 20:00 CEST (UTC +2), i.e 18:00 GMT and 13:00 PDT (UTC-7).
   12 posts; meanViews = 33; meanActions = 2; duration = 2.5 months
2. Monday 16:00 CEST (UTC +2), i.e. 14:00 GMT and 09:00 PDT (UTC-7).
   7 posts; meanViews = 35.5, meanActions = 2.6; duration = 1.5 months
3. Wednesday 16:00 CEST (UTC +2), i.e. 14:00 GMT and 09:00 PDT (UTC-7).
   ...

Spontanous posts:
1. Tuesday, 11 October 2016, 07:45 (UTC +2): Bounty question
views = 145; actions = 4 (highest ranked post!)
2. ...

T-tests

Between Monday 20:00 and Monday 16:00 no significant difference was found in the amount of view (t = -0.134, p = 0.895) nor in the ammount of actions (t = -0.664, p = 0.516), despite a (minimal) increase in members. It is not expected that increasing the sample size would be meaningful.