Re:Show HN: Email open rates are stupid.
by dangrossman on Tuesday November 01, 2011 @11:38PM
(#3185292)
I have no idea (didn't find out from a quick browse through the site), but the first thing that came to mind was an image in the mail, pointing to a server that slowly trickles a response without ever finishing so it can measure how long the other end is listening.
Re:Show HN: Email open rates are stupid.
by MrEnigma on Wednesday November 02, 2011 @12:17AM
(#3185399)
I worked at an ESP (Email Service Provider) a few years back. It may not be anymore, but it was a crazy environment back then of everyone trying to compete.
The major issue with this is going to be scaling it, with large ESP's doing 50-100 million messages a day, and 20-30% of those being opened, that's a lot of processes. This solution will probably be effective, mostly because it's going to a much smaller segment.
Re:Show HN: Email open rates are stupid.
by derwiki on Wednesday November 02, 2011 @06:05PM
(#3189154)
That's a neat trick. After hearing about Mirapoint, I built a simple proof of concept that uses time-delayed redirects to gauge how long a user has attempted to load an image:
Re:Show HN: Email open rates are stupid.
by derwiki on Wednesday November 02, 2011 @06:03PM
(#3189145)
If the sender is ReturnPath certified, Hotmail and Yahoo! automatically load images. Those two providers account for a large portion of email on the internet, so there's definitely value in this product.
Re:Show HN: Email open rates are stupid.
by Udo on Wednesday November 02, 2011 @10:32AM
(#3186852)
It's still bullshit. Most email clients don't display images by default. The notion that email reading behavior can be tracked with anything resembling reliability is a blatant lie.
Re:Show HN: Email open rates are stupid.
by FaceKicker on Tuesday November 01, 2011 @11:48PM
(#3185325)
Are images blocked on hotmail, yahoo, etc. too now? I've only used gmail for the past 5+ years.
Either way though, it wouldn't really produce a random sample, though maybe it would allow somewhat meaningful comparisons between separate email campaigns even if the absolute numbers aren't too meaningful.
Re:Show HN: Email open rates are stupid.
by rorrr on Wednesday November 02, 2011 @12:11AM
(#3185379)
If that's true, it's very dangerous. Imagine somebody finds an iphone image vulnerability, a virus based on it will spread across the whole world in a couple of hours.
Re:Show HN: Email open rates are stupid.
by phzbOx on Tuesday November 01, 2011 @11:43PM
(#3185308)
I'll be grumpy and say I'm surprised of how fast it reached first position on HN. However, I have to agree that there's a huge difference between deleting an email and opening it :p Great idea, good luck with it.
Re:Show HN: Email open rates are stupid.
by aarlo on Tuesday November 08, 2011 @02:10AM
(#3209819)
If they "click out" and move onto another message, they'll probably never read your email again. So for the marketer it's the same as a delete (and likely is an actual delete).
If they click on a link inside the message, that shouldn't get counted as a delete.
Feel free to ping me with more questions - [email protected]. Always looking to chat with others in the email world :)
I still can't imagine how they would detect deletion though.
The major issue with this is going to be scaling it, with large ESP's doing 50-100 million messages a day, and 20-30% of those being opened, that's a lot of processes. This solution will probably be effective, mostly because it's going to a much smaller segment.
3. When the connection to the endless GIF dies, call that the "reading" time.
4. If a "reading" time is under some threshold call it "deleted".
That's how I did it a few years ago anyway.
https://github.com/derwiki/redirect_tracker
Like most email metrics, it's lacking -- but this is still -a little- more information, and that goes a long way in email marketing.
You might be able to have some <img> link back to your server, but even those are blocked by default.
Either way though, it wouldn't really produce a random sample, though maybe it would allow somewhat meaningful comparisons between separate email campaigns even if the absolute numbers aren't too meaningful.
http://www.youtube.com/watch?v=gDW_Hj2K0wo
That would count as a delete?
If they click on a link inside the message, that shouldn't get counted as a delete.
Feel free to ping me with more questions - [email protected]. Always looking to chat with others in the email world :)