Subid-Referrer Analytics

• SubId:
  TPS + CustomTag
• All Searches:
  Number of searches recorded under that SubId
• JS Searches:
  Number of javascript enabled searches recorded under that SubId
• % JS:
  Percentage of total searches that had javascript enabled
• Total Clicks:
  Number of clicks recorded under that SubId
• CTR:
  Percentage of searches that generated a click
• % Searches From Bad IPs:
  Percentage of searches originating from an IP on Local's known offenders list
• % Unique Bad IPs:
  Ignoring multiple searches from the same IP, the percentage of IPs that are on the known offenders list
• Potential % Bad Traffic:
  Out of all the referrers that could come in under this SubId, this is the average percentage of known offending IPs

Notes:
A blank in the SubId column means that no SubId was provided from that source.
The totals at the bottom of the table change depending on what filters are chosen.

#1 - Suppose the search log shows SubId 6000 with:
3 hits from 1.1.1.1 
2 hits from 2.2.2.2  
and 1 hit from 3.3.3.3 
and supposing that 1.1.1.1 is a known offender.

This results in: 
50% - % Searches From Bad IPs, 
but only 
33% - % Unique Bad IPs.
Therefore, this large difference indicates there is a small number of offenders spamming us rather than generally bad traffic.
Thus, blocking the entire subId is not the best way to remove the offending traffic.


#2 - Suppose the search log shows SubId 6000 getting traffic from:
www.DellWolf.net
www2.omgrofl.org
and xinfire.media.com
and supposing that the percentage of bad traffic from those is 5%, 10% and 30% respectively

This results in: 
16.67% of the traffic that could come in from SubId 6000 is a known offender.
Comparing this to the traffic that actually did come in on SubId 6000 shows a large discrepancy.
Therefore the overall traffic that could arrive on SubId 6000 is better than what so far is arriving.
            

• Referrer:
  Http Referrer provided by our traffic, trimmed down to just the domain
• All Searches:
  Number of searches recorded under that Referrer
• JS Searches:
  Number of javascript enabled searches recorded under that Referrer
• % JS:
  Percentage of total searches that had javascript enabled
• Total Clicks:
  Number of clicks recorded under that Referrer
• CTR:
  Percentage of searches that generated a click
• % Searches From Bad IPs:
  Percentage of searches originating from an IP on Local's known offenders list
• % Unique Bad IPs:
  Ignoring multiple searches from the same IP, the percentage of IPs that are on the known offenders list
• Potential % Bad Traffic:
  Out of all the SubIds that could come in under this referrer, this is the average percentage of known offending IPs

Notes:
A blank in the Referrer column means that no referrer was provided from that source.
The totals at the bottom of the table change depending on what filters are chosen.

#1 - Suppose the search log shows Referrer www.DellWolf.net with:
1 hit from 1.1.1.1 
2 hits from 2.2.2.2  
and 3 hits from 3.3.3.3 
and supposing that 1.1.1.1 is a known offender.

This results in: 
16.67% - % Searches From Bad IPs, 
but  
33% - % Unique Bad IPs.
Therefore, this large difference indicates there is a large number of offenders but overall good traffic.
Thus, blocking the entire referrer is not the best way to remove the offending traffic.


#2 - Suppose the search log shows referrer www.DellWolf.net getting traffic from:
SubId 6000, 3200 and 4010
and supposing that the percentage of bad traffic from those is 30%, 40% and 50% respectively

This results in: 
40% of the traffic that could come in from referrer www.DellWolf.net is a known offender.
Compared to the traffic that actually did come in from www.DellWolf.net
Therefore the overall traffic that could come from www.DellWolf.net is worse than what is actually arriving.