Bot Cleaning
- Thread starter waypoint
- Start date
duffhome
Active Member
i think from mailwizz you can exclude and IP address from clicking.
I m wondering the same thing actually. I noticed lot of clicks from same IPs. when i looked it up, the IP belongs to Amazon. and i don't have anything to do with Amazon.
Those emails who clicked seems to belong to a real person.. i don't know whats that about!..could they be real persons using proxies? its like 1000 different clicks all from Amazon IP and other providers!!
For Spam Trap. A nasty business. Almost all major blacklist and anti-spam listing is based on spamtrap hits. (combined with DNS based, Authentication, downtime, query time, speed, volume.. ). you would not believe how many traps out there different types, domain belongs to different ISP and anti-spam services..
Spam trap are made to remain undetected. no special smtp response code, not special character or behavior, i been informed from different sources. that there are types of trap that can open email and click link so they can follow you...
to be honest i don't even know how to investigate that. unless those traps are managed by a real human. Like the original SpamHaus SBL, Where listing are created by individuals, but generally it takes hundreds or thousands of trap hits to get listed on SBL.
the best way i know to get rid completely of traps it by preforming a permission pass to you lists.
I would really love to hear what you and other mailers know and experienced in their mailing journey about bot clicks and traps
I m wondering the same thing actually. I noticed lot of clicks from same IPs. when i looked it up, the IP belongs to Amazon. and i don't have anything to do with Amazon.
Those emails who clicked seems to belong to a real person.. i don't know whats that about!..could they be real persons using proxies? its like 1000 different clicks all from Amazon IP and other providers!!
For Spam Trap. A nasty business. Almost all major blacklist and anti-spam listing is based on spamtrap hits. (combined with DNS based, Authentication, downtime, query time, speed, volume.. ). you would not believe how many traps out there different types, domain belongs to different ISP and anti-spam services..
Spam trap are made to remain undetected. no special smtp response code, not special character or behavior, i been informed from different sources. that there are types of trap that can open email and click link so they can follow you...
to be honest i don't even know how to investigate that. unless those traps are managed by a real human. Like the original SpamHaus SBL, Where listing are created by individuals, but generally it takes hundreds or thousands of trap hits to get listed on SBL.
the best way i know to get rid completely of traps it by preforming a permission pass to you lists.
Last edited:
frm.mwz
Well-Known Member
Re click bots, while the total of clicks and opens may be much higher due to them, if you focus on unique ones, the key info is the same. That leaves us with resource abuse stemming from those bots, for which there are blocking methods from certain levels onward, or some just use e.g. cloudflare.
Lots of clicks from the same IPs can come from spamfilters which do test e.g. also links.
Not sure if one can ever get rid of those bots, perhaps mitigation is the better strategy, since different providers may choose another firewall/spam/virus scanner/etc, and then another whole block of addresses would have to be excluded, and it seems better to have a general solution to be able to include all and solve the problem.
Re spam traps, there are millions of new ones coming up every day, so it is useless to even have a database of 100M. Better have clean, confirmed lists. The graphic @sendizo posted is generally right. There are quite a number of ways to go about this, e.g. segmenting by domain/mx to make sure you can fully focus on inboxing for each of those segments. Some say you can only run one permission pass, but if they signed up initially, and you send them transactional updates, then you have more chances to convert them into confirmed opt-ins, and even more, if they signed up for updates anyway (but initially without confirmation). Speed is out of the question for permission passes, they need to be done as gentle as possible (30-100/h).
Lots of clicks from the same IPs can come from spamfilters which do test e.g. also links.
Not sure if one can ever get rid of those bots, perhaps mitigation is the better strategy, since different providers may choose another firewall/spam/virus scanner/etc, and then another whole block of addresses would have to be excluded, and it seems better to have a general solution to be able to include all and solve the problem.
Re spam traps, there are millions of new ones coming up every day, so it is useless to even have a database of 100M. Better have clean, confirmed lists. The graphic @sendizo posted is generally right. There are quite a number of ways to go about this, e.g. segmenting by domain/mx to make sure you can fully focus on inboxing for each of those segments. Some say you can only run one permission pass, but if they signed up initially, and you send them transactional updates, then you have more chances to convert them into confirmed opt-ins, and even more, if they signed up for updates anyway (but initially without confirmation). Speed is out of the question for permission passes, they need to be done as gentle as possible (30-100/h).
waypoint
New Member
WoW all great feedback guys. So in the world of marketing emails permission pass sounds like a great idea @sendizo thanks for this input. @frm.mwz For the "general solution" I have heard some feedback recently regarding https links being more widely accepted due to the domain validation aspects for various ISP link validation techniques, any thoughts here?
frm.mwz
Well-Known Member
The more all data matches to your unique sending domain (no certificate vs generic cert vs your commercial cert with anon whois vs your commerical cert w/ your whois), the higher the reputation and the probability of inboxing.