Bots are not only used to scrape content and pricing, but also contribute to skewed site analytics. Both good and bad bots contribute to skewed analytics. If there are unexpected spikes in your analytics reports, chances are that these are from bot activities. Or, it could genuinely be one of your blog posts performing extremely well on that day. But, how can you be sure?
Luckily, there’s a way to identify and filter out certain bots from polluting your site metrics. Following are some steps that will help you identify bots from Google Analytics reports:
Traffic source: If the traffic source is direct, and that’s the main traffic driver for that particular day, it could be the bots
Bounce rate: If point #1 is true and the bounce rate is over 95% (generally, it’s 99% or so for bots), it could be bots
Avg. session duration: Just a few seconds? It could be bots.
Geo-location: If the geo-location is either not discernible or from all over the world, it could be bots (the whole world will not wake up one day, type your URL and visit your website).
Service provider: If most of the hits are from the same service provider, it could be bots.
In this case, since you were able to find the bot activity and tie it to a service provider, you can create an exclusion and filter this out. This will no longer create spikes and pollute your analytics. However, bots with malicious intents are not easy to identify and isolate. You may know from real-time analytics that there are so many ‘direct’ hits and the session duration is very short. The one thing you may not be able to do is pin-point the bot traffic source and isolate it by creating a filter.
Many bad bots come from referrals such as semalt.com, buttons-for-website.com, darodar.com, econom.co, and so on. All these websites are dodgy remarketing sites that are infested with advertisements and pop ups.
Sales, Marketing teams depend on Analytics from their web/mobile applications to learn more about the market and the audience that are using their service. With polluted analytics it is very difficult to differentiate traffic from unique human users and bots - making it very difficult to gauge the performance of marketing campaigns or tweak marketing strategies on the go to optimize the results. These bots take a while to become obvious as they slowly increase in number and eat up your bandwidth, slowing the web servers and using up resources.