5 ms·
I'm sending out a newsletter every now and then. A simple text newsletter that goes out to about 50,000 subscribers. I send it via a PHP script I wrote myself.
by xyby 12y ago
I'm sending out a newsletter every now and then. A simple text newsletter that goes out to about 50,000 subscribers. I send it via a PHP script I wrote myself. It can do a/b tests, bounce- and optout-handling. Every time I see commercial offerings like this one, I wonder: Would I gain anything by using them? They seem pretty expensive to me.
My selfmade solution also can do HTML mailings, but I never was able to measure any benefit of styled newsletters over text. So I usually use just text.
- kirkus 12y agoYeh you can better measure stats like open rate, CTR and deliverability. It's especially good for marketers and entrepreneurs that want an easy to use interface to manage their email marketing campaigns.
- xyby 12y agoAs far as I know, there is no good way to measure open rate. You can see if images in your html email are loaded. But that could also happen automatically or in a preview pane that is displayed by default. The reason why I shy away from measuring CTR is that I then cannot give people the real url. I like to directly give them the final url instead of some intermediate page. Maybe it's just me, but I prefer that. How do you measure deliverability?
- neeraga 12y agoYou're right, deliverability can be measured via pixel image itself. That is the same process that we follow at SendGlide.
- glomph 12y agoGoogle caches those images though. So it isn't accurate for gmail.
- aezell 12y agoOnly for repeat opens. The original open still accesses a particular URL specific to that recipient so the open is recorded even if Gmail hides IP, browser, OS, and other data via their proxy. If the user doesn't view the images, that open is never recorded anyway. So, the caching doesn't actually affect the open rate.
- glomph 12y agoThey cache them before the email is even read. They also cache images across gmail accounts.
- qu1mby 12y agoCould you use a url parameter on the end of the real url (e.g. example.com/page?source=email_[date]), and then use the history object to strip the param once they land on the url (modern browsers)?
- jvvlimme 12y agoDeliverability is going to be substantially better. IPs from commercial providers are trusted among ISPs and have a better chance not being flagged as spam.
- xyby 12y agoWhat would be a good way to measure that?
- jaredmiwilliams 12y agoYour bounce rate and reasons.
- xyby 12y agoA mail that goes into the spam folder does not bounce.
- thaumaturgy 12y agoYou're right for the most part, but I hope that will change soon. I've been working on a new spam filter for a while, it's in testing on my mail server for a handful of customers, and quite a bit of stuff from constantcontact, sendgrid, Amazon, and even Gmail are getting caught by it. The unfortunate thing is that every time there's a new trusted mailing service with a good reputation, spammers flock to it.
- aaronbasssett 12y agoNot in this instance. They use your AWS SES account, which you could just as easily use from your own script so deliverability will be equal.
- ThomPete 12y agoA rookie question if you don't mind me asking. How does your setup work? Have you thought about open-sourcing or selling the script? I am using MailChimp for weekendhacker but i always feel like it's way too bloated for my needs.