[cma-l] On-line Listener Figures

[bruce]

>
> My question is a rather simple one. How much faith are people putting
> into these figures? I have down an average number of listeners per
> month at 45,000 This seems excessively high and I wonder if the
> program is not taking into account repeat visits.
>
> I'm interested to hear peoples opinions.


Hi Chris,


As Bill says, the figures in the web stats are accurate - in that
they accurately represent what's in the server logs, and so represent
the number of accesses to your stream.

However, that's the easy question to answer. The difficult one is
"so what does this mean?"

Within the stats for each month, you'll find an entry "total listeners,
except 404". That's the total number of "hits" on the stream. Many
of these will be "real listeners". Others will be web crawlers such
as google etc.

There's also a stat ""total unique sites". That's an interesting one - it's
the number of unique destination IP addresses - ie the number of "listeners".


Now have a look at the "Listeners/Duration" graph. There will probably be a
huge peak at the left - lots of listeners who didn't stay long. Chances are
that these are the googles, and "casual browsers".

For my station, 26% of listeners stayed for less than 2 mins and a
further 12% for 2-3.. My gut feeling is that we shouldn't include
these numbers when trying to estimate how many real listeners we
have.

I've just noticed something - the numbers don't appear to tally
here. For example, when I look at our September summary, I get a
the total number of accesses and total unique access - but that
doesn't correspond with the tabulated figures which are broken down
by listening time - when you add them up you get a number somewhere
between the two.

Sorry if that's not clear, but there's no point in me posting real
numbers here. Bill, I'll send you my figures.

Bruce.


-- 

Bruce Rodger
Celtic Music Radio
bruce at celticmusicradio.net