All of us that use Google Analytics are to one extent or another blighted by bot and spam referral traffic. For those of us which rely heavily on data for business and marketing decisions its a real nuisance.
Whats the problem with bot traffic & spam referrals in Google Analytics?
- It can seriously reduce reported conversion rate
- It can negatively impact reported bounce rate
- It can over-inflate reported traffic volumes
In short it makes analytics data unreliable. It’s particularly frustrating when comparing data for periods of time as inevitably the comparisons can’t be trusted if you suspect data in 1 period is tainted.
How can you identify bot traffic & spam referrals in Google Analytics?
There’s a number of simple ways to identify this troublesome traffic:
- In the traffic sources report keep an eye on your ‘(direct) / (none)’ traffic as this is the most common traffic source that bots will show up in. If you notice sharp increases in this traffic source it could be a sign of increased bad bot traffic.
- In the landing page reports look out for a sharp decrease in Avg. Session Duration for specific pages. You might notice a lot of page visits with 0 time on page. This is a strong signal that traffic was not from a human.
- Similar to above look out for pages which have significantly increased or even 100% bounce rate.
- Look in your traffic sources report for referrers. Sources that always have 100% bounce rate and 0 seconds session duration. You can block these via a filter (more on that below).
How you can block bot traffic and referrer spam
- (Easy to do) The 1st thing to do is tick the Bot Filtering option in view settings. This will keep out some of the bot and troublesome spider traffic.
- (Easy to do) The next thing to do is to look at the referral traffic report and add the sites as an exclude filter. Either add the whole site name or if there are common words in the name just add the words (see below).
- (Trickier to implement) If you have access to your web logs and can identify pesky web bots or visits by IP you can block those bots and IP’s in your htaccess file with the help of this tool.
- (Trickier to implement) Because bots don’t always use the same IP address a more effective way of identifying non human traffic is to identify the User Agent. To do this follow this LunaMetrics guide. When you have done that the next job is to identify user agents with 100% bounce rate and 0 avg session duration by looking at the traffic sources report and secondary dimension ‘user agent’ (see below). IMPORTANT STEP – Make a list of such user agents and then in the same Analytics report enter those user agents in the advanced filter to make 100% sure they do not pass any valid traffic to your site. You do not want to block any genuine traffic from your stats. Once you have confirmed that you can set up an exclude filter in the view settings for those user agents (see below).