I’ve had a question for some time which I can’t find an answer to, which you might be able to help with regarding Keyword site statistics.

Am I right in thinking that if a keyword search features my site, it doesn’t necessarily mean it has been visited by a real person? It just means I’m in a list supplied by Google or whatever.

Interesting that you favour unique visitors as a measure of the “human audience”. With people potentially connecting through several devices, I tend to think of it as “unique devices”. Also, the “unique” bit means it only makes sense over a given time period. Are we interested in uniques per month, per quarter, per year etc? I guess it varies for different services.

One thing I would mention is that, while client-side analytics may not report on Javascript-disabled browsers, server-side analytics tend to report all people from the same company as one person (due to accessing via the same IP address and user agent). Server-side data also needs to be filtered with the latest robots and spiders exclusion list while client-side has a more natural filter built in.

Lastly, it might be worth mentioning combined metrics such as pages per visit or visits per visitor because they tend bring these sometimes huge numbers into something more manageable. It’s amazing what dividing one by the other can do!

And where were you when we were beta testing Analytics?

Actually, if you haven’t checked it out yet, I’d love to get you in there. Perhaps sweeten the pot with three free months in exchange for feedback/review? If you’re interested, just give me a holler.

I actually think maybe my tumblr bounce rate isn’t so bad given that my updates are also on the homepage, or is there a time limit on it? Is it people who leave within 10 seconds of landing on the page or something like that?

Anyway good thorough introduction for people like me.

Very handy, this, as a reminder for myself and somewhere to send others for a complete overview.

Good one.

One of the interesting things I’ve found with Google Analytics is that a fair few people don’t realise how much it can do, so they over-discount it. ‘It’s free so it can’t be that good’ – Google’s so big they think there must be some sort of (bad) hidden motive.

Couple of add points: you can share views and that can be very useful for looking at why or whether a similar page is more/less successful (I’m going to be looking at promoting this for local government); through adding code to URLS you can track clicks in places like email newsletters, email clicks on pages, PDF clicks.

The SocMed stats stuff is also useful btw (forgot to mention).

@Paul C: fair point I look at a lot of this through the lens of Google Analytics which is undeniably awesome for the price. But there are plenty of other packages worth exploring once you want to do more sophisticated things like track groups of pages, run more customised reports or get stats on internally-hosted tools.

@Paul J: interesting – I’ve updated above.