Inaccurate open tracking

[twisted1919]

Could you explain why not above should not work to get accurate open data?

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

This user agent and IP field should that of the recipient.

Not always, see my above example. SG can't report a correct ip address, mine in this case, if it will only see google's ip address.

 

[mobibeasts]

​

@twisted1919
Obviously,
It is possible some anti-spam filters or recipient servers might employ spam filters or software to verify that an email, is in fact, safe for a human to open.

So based on your confirmation,
currently, mailwizz does not inherently decipher between legitimate and illegitimate opens because the signal is essentially the same for a user and a bot. However, there is a behavior you can use to try and isolate false positives caused by bots or filters.

However,
You can potentially use the useragent field that would be provided for open events and passed in the email activity feed or event webhook to uncover more information regarding where these events originated.

User-agent strings should contain information that look like real-world combinations of devices you know. Web crawlers, mail filters or other pieces of software can be identified using the user-agent, potentially programmatically.

 

[twisted1919]

So based on your confirmation,
currently, mailwizz does not inherently decipher between legitimate and illegitimate opens because the signal is essentially the same for a user and a bot.

This is correct, exactly!

User-agent strings should contain information that look like real-world combinations of devices you know

Sure, but 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.
See:

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

 

[nadworks]

Just throwing the latest changes with Apple's iOS 15 release into the mix, which will further mess-up the tracking landscape...




We'll still need improved open and engagement tracking. But we will also need to keep an eye on these changes and be flexible enough to change our approach as it might become a creative / communication / education challenge above all else.

 

[creativeandrew]

Just throwing the latest changes with Apple's iOS 15 release into the mix, which will further mess-up the tracking landscape...

View attachment 13987
View attachment 13988

We'll still need improved open and engagement tracking. But we will also need to keep an eye on these changes and be flexible enough to change our approach as it might become a creative / communication / education challenge above all else.

Exactly! I don't think Opening Rates will be useful in the near future and even with a fix, they will never be accurate. We will need to work only with Click Rate, am I right?

 

[nadworks]

Exactly! I don't think Opening Rates will be useful in the near future and even with a fix, they will never be accurate. We will need to work only with Click Rate, am I right?

Nope, sorry. I completely disagree. They will be challenged, but the jury is still out about the open metric. And I consider it essential to work on getting it close to the same level as other ESPs have managed to.

It'll be absolutely vital for Email providers (Gmail, MSN, Yahoo etc.) to keep their customers' inboxes clean, so these giants will somewhat be on the side of tracking and transparency. Helping email marketers like us to suppress recipients who are goneaways or have stopped engaging with our brand is incredibly important to keep email traffic manageable. You would basically be forced to water down your messaging just to desperately include a click-through, even when none is needed.

I feel education and clarification is much more important.

 

[creativeandrew]

Nope, sorry. I completely disagree. They will be challenged, but the jury is still out about the open metric. And I consider it essential to work on getting it close to the same level as other ESPs have managed to.

It'll be absolutely vital for Email providers (Gmail, MSN, Yahoo etc.) to keep their customers' inboxes clean, so these giants will somewhat be on the side of tracking and transparency. Helping email marketers like us to suppress recipients who are goneaways or have stopped engaging with our brand is incredibly important to keep email traffic manageable. You would basically be forced to water down your messaging just to desperately include a click-through, even when none is needed.

I feel education and clarification is much more important.

I understand your point. But unless I have understood it incorrectly, once the Apple Mail update is in effect we won't be able to track down openings as all emails will be opened by a proxy by default due to the new privacy regulation.

 

[nadworks]

I understand your point. But unless I have understood it incorrectly, once the Apple Mail update is in effect we won't be able to track down openings as all emails will be opened by a proxy by default due to the new privacy regulation.

Wrong. It has nothing to do with proxy and caching, and so far it will only affect emails opened with the iOS 15 native email app for people who actively opt-out of tracking.
Hence my thoughts around education and transparency from both ESPs and email marketers.

 

[Ashish Shukla]

Anyone have solution for that?

In future other ISP may also include same privacy. @twisted1919 Any plan on your genius mind ?

 

[Jason Thompson]

Hi @twisted1919 i've spent a lot of time researching this topic and came across an interesting article that could be a solution to this google image proxy issue.

From what i've researched, large companies that send bulk emails use the "bgsound" method instead of image tracking pixels.

Here is an example of Facebook tracking with "bgsound"
<bgsound src=3D"http://www.facebook.com/email_open_log_pic.php?mid=3D51178b2G56e57c6dG1f415bcG0&s=3Da" volume=3D"-10000"/></span>

Do you think this would work with mailwizz?

Here is the article: https://www.zdnet.com/article/facebook-uses-bgsound-to-see-if-you-opened-an-e-mail/

 

[twisted1919]

@Jason Thompson - I just opened an email from hubspot and one from mailchimp, they both use regular tracking pixels.

Do you think this would work with mailwizz?

We can run tests and see how it works, sure, I'll get back to you next week with some answers

 

[Soleil34]

Hi everyone.

I'm a new customer and user of the software which I installed today and currently testing.

I have version 2.0.21

I have been experiencing some weird things with the tracking of opens and several searches of the forum brought me to this thread which I have just finished reading.

For info - All my emails go through Amazon SES API deliver server. That is all setup correctly and deliverability is fine. I have 10/10 score at mail-tester. All emails are getting through.

I have set up a seed list of 3 email subscribers:

Subscriber 1: A gmail address
Subscriber 2: A gmail address
Subscriber 3: A yahoo address



If I close all tabs to gmail and yahoo mail and then send a campaign here are the results:

- immediately (before I open any of the emails) the opens stats for the campaign record 1 open for subscriber 1 (gmail) - I see that being discussed in this thread. But I also have another problem with gmail:

-If I open "subscriber 2" gmail account (via webmail) and open the email (in the "primary" tab FWIW), the open is NOT registered in mailwizz. - I have checked the raw message and the tracking pixel is in the raw message. If I copy paste the URL to the tracking pixel into a browser, mailwizz does then register the open.

-If I open "subscriber 1" gmail account (via webmail) and open the email (in the "promotions" tab FWIW), the open is NOT registered in mailwizz. I have checked the raw message and the tracking pixel is in the raw message. If I copy paste the URL to the tracking pixel into a browser, mailwizz does then register the open.

-If I open "Subscriber 3" (yahoo) email account (via webmail), the open is correctly registered in mailwizz. Every time I open the email, it's registered correctly.

So to summarise:

I send a campaign to 2 gmail accounts and 1 yahoo account. Immediately (before opening anything), one of the gmail accounts registers a false positive.

When I manually open either of the gmail messages, neither of them register an open when using the gmail webmail interface. (Anyone else getting this?!)

Yahoo mail works fine.


I understand the gmail immediate false positive is a known issue. But what about the case of gmail opens by a human not being tracked? Is that a known issue too?

Cheers.

 

[daleji]

@Jason Thompson - I just opened an email from hubspot and one from mailchimp, they both use regular tracking pixels.


We can run tests and see how it works, sure, I'll get back to you next week with some answers

Hello,

If I have to test, where should I change?

thanks.

 

[WWJ2]

Hi,

Would this be a possible solution?

Create a unique open pixel for each email sent. When it gets hit, it gets decrypted and stores the user id and campaign Id, so it didn't matter what IP it comes from.