Various articles have appeared recently discussing BOT traffic and how this type of traffic is a great danger to the integrity of online advertising. And it is true that BOT traffic is a real problem with an estimated 10-20% of all traffic being BOT (11% of online and 23% of video according to the WSJ).
This has resulted in a slew of solutions that promise to “clean up” the traffic by measuring and filtering out any BOT traffic. Of course these solutions come at a cost and to make sure you buy most will make you test your traffic and find as much as 60% BOT (especially if you run a small website – You are prime target of these people). However, while these solutions claim to do a good job of identifying and cleaning up BOT traffic, most are vastly over shooting and identify and eliminate what they identify as “suspicious” traffic but that in reality is perfectly legitimate traffic.
Take for example the case of a referrer that is not passed by a source of traffic. Most BOT filters will identify this traffic as BOT but in reality this traffic in many cases is totally legitimate. Many networks for example do not pass referrers and “domain” traffic is more often than not masked but the traffic is legitimate. As a result, we’ve seen very solid conversion traffic labeled as BOT. The reason is that more often than not the algorithm designed to identify BOT traffic is not capable of distinguishing between a source of valid traffic and a source of BOT traffic.
On the other hand we’ve seen so called IAS rated traffic that guarantees 100% human (with the restrictions mentioned above) being tested and proven to be fraudulent. The reason is simple – some traffic providers cheat. And since many web owners do not have the IAS test code on their site, they may be paying for traffic that is BOT and not realizing it, until the advertisers on their site ban them. Additionally, a simple fact is that while some BOT traffic is easy to identify, most people who send BOT traffic are not dumb and are adapting.
So what is the solution?
First and foremost get your traffic from reputable networks or sources that are testing the traffic that they provide and are able to detect BOT or non performing traffic before it comes to you. A CPM for a certain type of traffic or a click have usually known cost brackets so any traffic that is offered at a fraction of what most would sale it for has to be viewed with caution.
Second, understanding what is your traffic type. Is it search, pop, domains, etc.? This will give you the ability to assess the traffic based on referrer being passed or not and as a result if referrer or IPs mismatch are legitimate for example.
Thirdly, finding and implementing simple solutions to test your traffic to insure that it is human. These include simple testing like CAPTCHA on part of the traffic that you get (2-3% is enough with at min 25% passing). Other solutions include blocking traffic from known BOT (published by IAB if you are a paying member), blocking IPs that generate more than one click per day on an ad, implementing honeypot links (click traps – these can even be implemented by yourself via a simple 1 pixel image and link), measuring mouse movement, off-screen clicks, etc. These are great options for rapidly identifying BOT traffic and, often, they are available free by ISPs.
Finally test your traffic continually and no matter what the quality. This is the best way to insure that the traffic that you get is clean day in and day out.
Bottom line is that, in our view, the best solution to addressing and managing BOT is to be aware of where your traffic comes from and test it and continue testing no matter how reputable the source of traffic may be. People who send BOT are smart and they are always going to find a way to fool algorithm no matter how much we think we are in control.
Just as a side note, here is a diagram that I saw from Solvemedia back in 2012 entitled “The BOT Stop Here” (click on the image to enlarge it). It has a great illustration of estimates that show anything from 4% to 31% BOT traffic in advertising depending on who you talk to and some down to earth recommendations. Bottom line it is your $ that is at play so make sure you are careful.
Additional resources on this issue:
- Wall Street Journal – http://www.wsj.com/
- Website Magazine – http://www.websitemagazine.com/
- IAB List of BOT – http://www.iab.net/
- Malware Domain List – http://www.malwaredomains.com/
If you have any questions, let us know by contacting one of us via firstname.lastname@example.org