Becoming Conversant in Google Analytics: Timing Posts

Are Google Analytics necessary for beginning bloggers?   Here’s one of the cool aspects of Google Analytics: its capability to show us activity on our blogs by time of day.   This helps in figuring out what to post, when!

Visualizing User Data

One of the biggest problems is making sense of the visualization of the data.  Here’s a run down of my current, VERY small blog’s analytics as provided by Google.

Screen Shot 2018-02-22 at 7.20.09 PM

Oookay.  Google Analytics logged 78 users in the last week, who had 85 sessions. These good numbers are going up. (nice, you can tell by the little arrow)  and the bad numbers (bounce) is down.  I have no idea what those two lines mean.  What is the dotted line about?  What is the solid line measuring?

Solid Line vs Dotted Line. You have to hover over the data with your mouse cursor.  The faint dotted blue line is last week’s data.  The solid blue line is this week’s data. The helps to see how the two weeks compare. Stats for the last seven days provides a different view:

Screen Shot 2018-02-22 at 7.29.25 PM

In this way, Google Analytics can be more reassuring (and LESS reassuring) all at the same time.  WordPress shows many fewer visitors (33 as compared to 78) but it’s already been noted that WordPress stats are just not the same as Google Analytics. Those who use the WordPress reader do not entirely register as having visited my site in the WordPress “Jetpack” statistics. 

The Bounce Rate and Session Duration

The “bounce rate” is the number of times people only read ONE PAGE of your blog.   If you’re trying to get people to stick around and BUY something, then, you want to reduce the bounce rate.  My blog is fairly geared to having people stop by for just the one page. WordPress Reader encourages this and it’s the main traffic driver for me.

The odd thing is session duration.  22 seconds is the average?!!? (Do people really only stay on my page long enough to read the first sentence?) I wondered if this has anything to do with how the WordPress Readers register in Google Analytics. They may register as having connected to the page–which is better than not registering as having visited the page at all.  This is what happens in WordPress stats–those readers who only read through the reader are not counted in WordPress stats, and it may follow that the the duration of their visits may not be counted correctly in Google Analytics, as well.  I’m not sure what’s going on there and neither are WordPress’s Happiness Engineers I’ve consulted.  We’d need to get together and do some experiments, I think, to determine the difference between GA and WordPress analytics.

Users by Time of Day

Screen Shot 2018-02-23 at 8.11.34 AM

This chart from GA is useful in that it shows that get lots of readers very early in the mornings (from the East Coast perspective. I ‘m fairly sure it’s East Coast, but as I don’t recall SETTING my time zone, I’m unsure. Of course The Google Knows All, so I’m guessing it knows my time zone.)

My audience apparently are morning people. 🙂  Also note that they are consistently “night people,” too.  There’s usually someone up around then, or I have this small group of night owls slash vampires who work on blogging things late in the night.  Daytime is very sporadic (from the East Coast perspective).   Knowing when our readerships are alive and kicking, er, reading, well that could be useful.

WordPress Statistics don’t give you this degree of specific information.  Instead, they make averages over the entire year (I believe) and provide you with a very different metric.  Here’s what WordPress Stats provide:

Users by Day from WordPress.Com

Screen Shot 2018-02-23 at 8.23.14 AM.png

Averages over the entire year are not as helpful for very small blogs like mine.  I can shift Google Analytics to look at the average number of views by hour in the last seven days, in the last 30 days, the last quarter–oh, and I can set custom ranges as well!

However, the annual, very specific information provided here gives me useful information–number of likes, for example–that are unavailable on “GA” (Google Analytics).

The Problem with Averages

The problem with averages is that the more data you give them in a dynamic, cyclical event, the less useful the average becomes.  If you were measuring how many presents children get every day from their parents, the numbers would be ones –and added to that all the numbers they get at Christmas — and then divide that by 365 days, you’d think children everywhere in the West were getting several gifts every day.  This would be incorrect. Most kids get the lion’s share of their presents at Christmas.  The big bumps would skew the average.  It’s important to look at data that varies by time in useful chunks.  This is why Google Analytics data may have greater utility for timing.  It’s not just that it provides you information by hour, by day and by week–it also lets you look at those rates by different time intervals.

Checking Our Data By Looking at Different Time Intervals

Screen Shot 2018-02-23 at 8.11.13 AM.png

In this view, we can see that the two time intervals with heaviest readership–at 9 and 10 am on Wednesday–remain constant when we look at the entire 30 day period.  Still, late night posts on Sunday also get significant traction over all, information that’s washed out if you only look at the last seven days.

The two dark blue top reads are an artifact of this week–they’re like Christmas.   They look very important for this week’s data, and when you look at the last 30 days, they’re still important.  But late night Sundays is also important, consistently, over the last 30 days, and there’s a great deal of variability in the data.

 When we look at the data for the last 90 days,  we see that this pattern tends to hold. 

Screen Shot 2018-02-23 at 8.52.11 AM.png

So is there a real “sweet” spot for posting?  I think the data here is far too variable to say just yet.  Feb 12 (last week) was a banner week for this blog.  What did I do?

I wrote a provocative and fun opinion article that got lots of reads. I backlinked them to the Daily Prompt.   That’s my hint for getting great stats — be active in the WordPress Community!

(Conversant is the Daily Prompt Word of the Day.)

So today, I’m trying it again to see if the impact of Daily Prompt. 🙂   I’ll update this later with the stats.

One More Thing

Consider one more issue: when are you WRITING these posts?   My map of readers’ times shown here is also a darned good map of WHEN I PUBLISH posts:  usually around 8 am and around 6 pm.  On Saturdays and Sundays, it is ofter around 2pm.  About five minutes after I publish, a tweet goes out and a post goes up in the WordPress Reader.  That has to account for some of activity.

However, Monday tends to be a very good day even according to GA.  And that gibes perfectly with the information from WordPress Analytics–which also says Mondays, at 1pm.   So, don’t over think it.  Google Analytics is fun, but still not a fundamental requirement. 

More soon.  Thanks for reading!  ~ Lola

Google Analytics tell us quite different things about our posts.  Here's how Google Analytics can help us get a better understanding of when our readers are engaging our content.

Author: Lola

Recovering academic, real-life, honest to cornflakes anthropologist (Ph.D. and fieldwork and everything), tech-head and social media researcher.

One thought on “Becoming Conversant in Google Analytics: Timing Posts

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.