Google Analytics for a long time has allowed us to exclude certain tranches of traffic from ever appearing in our reports. What's important of course is making sure you are excluding everything you need to, not just the standard stuff. Standard exclusion - your office traffic

An example of this would be to exclude your internal office traffic. You know the kind; marketing bods updating the website, bosses checking it's how they want it to look and so on. How do we do this?

First, login to your profile. I'm assuming that you're using the new version of the Google Analytics interface.

  1. Click the cog wheel to access your settings
  2. Click on the filters tab
  3. Click on +New Filter
  4. Fill it in as per the screenshot. Exclude, traffic from IP address, Equal toGoogle Analytics exclude traffic from office
  5. And now just fill in your IP address (IT should be able to tell you) or go to Google and search "what is my ip address" and it will tell you what to fill in. Please note if you're at home it's not worth doing this most of the time because home IP addresses tend to change every few days.
  6. Now go to your standard reports to add an annotation on today's date, so that you know when you added this exclusion Annotations in Google Analytics

Now, is that it? No it isn't. This is just a pretty basic step to exclude your office IP address traffic from your profiles - both internal traffic and your agency's traffic - giving you more accurate reports help you make better decisions and give you better averages. But there's some other stuff you need to be aware of.

The sneaky stuff

The really sneaky things are the items that you might not get told about. We were talking with one of our clients recently and saw that their average bounce rates had shot up, but traffic was still about the same, as were their online sales. How is this possible?

A bit of digging later revealed the sneaky stuff. The internal IT department had started monitoring the website. Normally this isn't a problem as a lot of services use a 'ping' which just checks the server, it doesn't access the website. What we saw here though, was a tool that was checking and downloading the home page of the site every single minute.

Now for those who understand bounce rate, this basically means I landed and left immediately (visiting exactly one page) - it had a 100% bounce rate and was triggering hundreds of visits a day. This was significant and having a large effect on the bounce rate.

If you go into Audience > Technology > Network this is where we spotted a lump of traffic coming from a particular hostname and so excluded the traffic in the same way as before.100% bounce rate, 1 page per visit, zero time on site. Massive efect on the averages.

Hostname producing poor Google analytics averages

The lesson here is to never ignore a shift in a metric that is significant. There will be a root cause of this shift that needs to be found. What this particular one was doing was not only inflating actual real traffic sources, but had concealed a significant drop in traffic from a very important referrer to the site - something we can now action to get this traffic back up.

Beware the sneakies!

1 Comment