An Important Update

Dear Followers Of This Blog ...

If you did not use a Blogger / Google account when you Followed this blog, years ago, you are probably not Following now . During the past...

Tuesday, March 31, 2015

Stats Displays Are Significant, In Their Own Context

Some blog owners want significant details, from Stats.
How do I get all posts listed, with pageview counts?
or
How do I get a specific time period, from last month?
or
How do I filter out the bogus counts?
Not everybody observes that Stats gives us specific details, based upon what's generally useful, or what is possible to keep, easily.

Some details are not provided, because Stats provides details that are significant, in the right context. Other details are not provided, because they are not retained, from day to day (week to week, month to month, or year to year).

Some blog owners carefully check their Stats logs, daily, and copy the contents into spreadsheets, and add up all countries. Or, they add up the last weeks totals, for any one post.

They may ignore the fact that Stats provides only the top 10 entries, for any time period - and that the top 10 will change - from day to day, week to week, and month to all time.

They may also ignore the fact that Stats time periods don't, necessarily, add up into other time periods.

  • "Now" is simply the last 2 hours, at any time.
  • "Day" is simply the last 24 hours - and "Today" may be anywhere from the last 1 minute (at 00:01), to 1440 minutes (24:00).
  • "Week" is simply the last 7 days. The current "day" may, or may not, consist of the most recent 24 hours, in the "Week"
  • "Month" is simply the last 30 "days". This won't be the last 4 "weeks", exactly.
  • "All time" is what it says.

Recalculations and resets won't occur according to local clock / calendars.

At any time of the day, week, or month, this week, month, or year may include some portion of another day, week, or month. The statistics for each day, week, month, or year may or may not be reset at the same time as the day, week, or month that just ended.

"Today" does not end precisely at midnight, for 23 / 24 of the worlds population - nor does it end at midnight for the same population, 100% of the time. If you think about it, the displayed "Week" may not be exactly the last 168 hours, either. Nor is the displayed "Month" going to be the last 720 (or 744) hours (does this month have 30, or 31, days?). Each time period is simply significant, in its own context.

All Stats displays are limited to the 10 entries reflecting the highest pageviews.

Any blog of any size will have more than 10 posts - and there will always be some posts that are more interesting to the readers, and get more traffic. Instead of using displays that could be infinite in size (how many posts would be part of any given blog?), they display the 10 most significant.

All Stats displays include pageview counts affected by referer spam.

Some "pageviews" may simply be referer spam, and may not indicate actual traffic. Referer spam follows no observable pattern, it simply appears as noise.

Posts with a low (or zero) level of organic activity are just as likely to attract referer spam, as posts with a high level. And blogs with low organic activity will show devious / low level referer spam more naturally.

Referer spam causes spikes in Stats logs, both when it is attempted, by the spammers - and when it is removed, by Google.

Spikes in Stats logs are more interesting, for posts with a low level of genuine visitor activity - and for blogs, in general, with low levels of genuine visitor activity. Like worrying about ones own pageviews, worrying about referer spam is more important to newer blog owners.

Fluctuations are more significant for blogs with low reader activity.

Most experienced blog owners don't worry about their own activity, or about imaginary activity - they focus on genuine visitor activity. Similarly, more experienced blog owners focus on the posts with more visitor activity.

You cannot add counts for the most 10 significant posts (pages), and compare to any traffic source, or audience. A spike in the reader activity for a given page may, or may not, correspond to a similar spike in the audience. You cannot correlate posts to audience, and deduce that any one post is more interesting to readers in a given country.

What you can do is examine the 10 most significant countries, during the last week - or the 10 most significant posts, during the last month. Neat month, you can compare the 10 most significant posts, against totals retained from this month.

You cannot extract this months totals, from 12 months ago, for comparison - nor can you extract all posts, for today, to add up. Nor will adding anything up tell you anything useful.

Accept a Stats component or display, for what it is.

Dude, hit me with a comment!

Carlos Molina said...

Hey. That was an interesting reading. Do you know how to filter the top 10 and filter pages by keywords or something? Thanks in advance.

Jadav Hiren said...

Thanks u very much.