Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open rates are inaccurate #1025

Closed
MaximilianKohler opened this issue Nov 15, 2022 · 5 comments
Closed

Open rates are inaccurate #1025

MaximilianKohler opened this issue Nov 15, 2022 · 5 comments
Labels
question Further information is requested

Comments

@MaximilianKohler
Copy link
Contributor

MaximilianKohler commented Nov 15, 2022

Version:

  • listmonk: latest
  • OS: I'm on Windows 10 x64, Listmonk is installed on latest version of Ubuntu server on AWS EC2.

Description of the bug and steps to reproduce:
As soon as I send a campaign, the open rate hits 50% within a minute or two. If working "normally" the open rate should start off at 10% or lower, and throughout the next 24 hours rise to around 60-70% as people open their emails.

I've been recently testing lots of other email programs such as Mailerlite, Getresponse, Convertkit, Sender, SendPulse, Mergo, and more. They all work "normally". But I experienced this same problem with Sendy (which also sends through AWS SES), which is what lead me to want to see if Listmonk has the same issue (it does).

I also read about this on the Mailwizz (another SES UI) forums:

And it sounds like they may have fixed it by blacklisting some bot IPs that automatically scan emails & links. I have yet to test Mailwizz to confirm though.


EDIT: and to clarify, I'm referring to the "more accurate" open rate & percentage listed on the campaign analytics page, not the less accurate "views" listed on the "all campaigns" page.


More info on how it works:

Thankfully it only counts each open once (if one person opens their email 5x to read it again it only counts it once), so you can subtract the starting opens to get a decent idea of the total. Unfortunately the analytics page becomes totally useless. And if you have a "view in browser" link, it'll count that each time its loaded (plus the original loading of images in the recipient's email). It does not count "unsubscribe" links; neither at the bottom nor in the header. IPs don't seem relevant; if you send to two different emails, and open each on the same PC, it counts as 2 opens.

@MaximilianKohler MaximilianKohler added the bug Something isn't working label Nov 15, 2022
@knadh knadh added question Further information is requested and removed bug Something isn't working labels Nov 18, 2022
@knadh
Copy link
Owner

knadh commented Nov 18, 2022

hm, I'm unsure what the issue here is. A view is registered when the view pixel loads. If bots are rendering messages and loading the image (not sure if that's a common scenario), it'll be counted as a legitimate hit.

@MaximilianKohler
Copy link
Contributor Author

If bots are rendering messages and loading the image (not sure if that's a common scenario), it'll be counted as a legitimate hit.

So can you try blocking known bots? For some reason, the vast majority of email marketing providers have "normal open rates" and this is only an issue with the few SES UI options.

Maybe it's because of the top comment here? https://forum.mailwizz.com/threads/inaccurate-open-tracking.7557/page-2

Mainly because if you're sending through smtp, you can't collect this type of info.

Maybe this "bgsound" comment is useful? https://forum.mailwizz.com/threads/inaccurate-open-tracking.7557/page-2#post-47358

@knadh
Copy link
Owner

knadh commented Nov 19, 2022

Not feasible to integrate bot blocking in listmonk. It's outside the scope of the project. You can run some sort of a web application firewall (WAF) or proxy in front of listmonk that does this.

@ajoslin
Copy link

ajoslin commented Dec 18, 2022

So a firewall rule on Cloudflare with block known bots could work?

https://developers.cloudflare.com/firewall/known-issues-and-faq/#caution-about-potentially-blocking-bots

🤔

@MaximilianKohler
Copy link
Contributor Author

So a firewall rule on Cloudflare with block known bots could work?

Based on the discussions I linked to that doesn't seem like a valid option. Quote:

in my case, when i opened the email personally, not just that I got google's IP address, but i also got google's image proxy user agent.

So while I did the open, a legit open, there is no indication here that I did it, it's just the google bot.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
question Further information is requested
Projects
None yet
Development

No branches or pull requests

3 participants