Bot traffic

As already mentioned the IAB list only works for bots that are listed on it, andnin my experience that's a very small percent of bot traffic hitting our site. You can also add custom bot rules based on user agents or ip addresses/ranges. But that might not be enough either. This is a good option for some things,  ut for general bots, because companies can buy ranges of IPs these can change frequently and can be hard to maintain. 

What we've always done is make a bot filtering segment with various conditions to identify visits from bots and then create a virtual report suite with the bot filtering segment applied to it. To create this segment we often use combi actions of country/region, browser, referrertype, domain, and screen size. We use metrics like total visits, bounce rate, conversion rate, and a few other engagement Metrics (since bots don't click on links, add products to cart, or buy stuff, if a visit is engaged it's probably not a bot). 

It will likely take a bit of work to determine what traffic is coming from bots, but as you get more experience, it'll get a bit easier to notice. But for your specific case, I would start with using country for Singapore, look at visits and conversion rate, break it down by the typed/bookmarked referrer type, and then keep breaking down by either domain, browser, or other dimensions until you find segments of traffic with high visits and no conversion or other engagements. Those will be your bots. Then make a segment to exclude them.

In addition to what the others have said, some bad bots do operate out of Asian (or seemingly Asian) countries… We had an issue a few weeks ago coming from Vietnam.

We debating on adding custom rules to our Bot rules, but we ended up instead working with our DevOps team and blocked it completely at the CDN level… it wasn’t legitimate traffic, odds are it was scrapping our sites for content, so the decision was made to just prevent it fully from accessing the site… 

In our case, we could only see a clear “bot” traffic when we looked at the data at minute granularity. Under normal reporting, the data didn’t outright “scream” bot traffic.. but at super small time increments, we could see the traffic coming in for 1-2 mins at a time, every hour or 30 mins… the pattern was clearly not human.

Hi ​@SousaJo 
Please consider marking a best answer so it could help others who have a similar concern in the future.

Just an FYI, I was in a Deep Dive session on Bot Detection a few months ago, and the video was just posted (there were some unforeseen delays), anyway, there’s a lot of ideas and processes covered that you (and anyone else who comes upon this thread) might find helpful.