Botman <> AdKernel Configuration FAQs & Example

Home  >  Knowledge Base  >  Botman <> AdKernel Configuration FAQs & Example

Examples

 

Q1. What is Control Group of Clicks and Max Suspicious Rate. Explain with example.

A1. TQ = High

Firewall Mode is OFF

Max Suspicious Rate = 40%

Control Group Clicks = 2000

Control Group of Clicks: The AdKernel engine won't take decisions about limiting the requests to the feed till it gets 2000 clicks from a certain source. After getting 2000+ clicks it will start blocking the source which it's not reaching High TQ rate. 

Note that it won't block all requests at once, it still will allow some traffic in order to check relevant quality (in case it got better, the engine will allow more requests). 

Max Suspicious Rate: In this case, if the number of clicks from a certain source with TQ than required level won't be higher than 40%, the requests from this source won't be limited. 

Q2. If we set up TQ=Low,  the engine will not send any click to check.  What is the point of having TQ=LOW? And what would happen if TQ=LOW and Fire Mode is ON? 

TQ = Low can be used in case of direct campaigns. For example, you set Low TQ for Offers remote feed and allow all clicks. But for certain campaign, you can set TQ = High and for such campaign, clicks will be checked and will be filtered in case if their TQ won't be High. 

Q3. It was my understanding that a click will not be filtered out if Fire Mode is OFF.  However, it appears, clicks will still filtered out once it exceeds Max Suspicious rate. is this correct?

A3.. No. If the firewall mode is kept off then no clicks are sent to the fraud check system for scoring. Max Suspicious Rate and Control Group of Clicks options are used as a requests limiter. 

Q4. Control Group of Clicks:  Is it for the number of clicks for Pub’s Subid level?  or for an entire pub level? Does it reset every day or one time?

A4. Per subid. It's a cumulative value that works for all period of time. 

 

Q5. Suspicious Rate:  How do you calculate the rate?  Do you check daily or cumulative from the beginning?  As I understand, after it passes the % threshold, the source would be reduced  around 48 hours. How much does it get filtered out?  Until it reached below the threshold?

A5. Click ratio (amount of clicks with TQ lower than TQ level / all clicks). E.g. 40% means that we will allow requests from a certain source if the number of Low TQ clicks will be up to 40%. if the number of clicks will reach 41%, then the engine will start limiting the requests from this certain source. 

Q6. I am just starting. What is the best practice to observe on sampling?

A6. That would depend on volumes of course. To mention subjectively, the best practice would be to start with a higher % first and then decrease down for more or less old / known sources, more like for background sampling. 

Q7. Will my referrer change if I use a filter site on the supply side?

A7. Yes, the referrer will change to the filter site domain if its used


Hardik Gandhi