I assume he meant "Boils down to your [choice of] infrastructure", as in, none of the two is inherently better than the other, and your choice would depend entirely on what you prefer to use.
We've also been using Sendy and I agree, the code is far from perfect and we have spent countless hours either fixing some issues or adding much needed features, making it actually usable for SaaS (sentopia.net), will need to move away to something better in the near future, this maybe a good start!
Probably the worst code I've ever seen, copy and paste gone crazy, buggy as hell in tracking where it is up to, just nasty and I really wanted to like it. Refund was no problem though.
Probably the worst code I've ever seen, copy and paste gone crazy, buggy as hell in tracking where it is up to, just nasty and I really wanted to like it. Refund was no problem though.
I can't compare, but I can comment on Sendy. I use it to send about to a list of about 5k.
It has never sent to completion on its own, it always requires me to monitor the queue so I can hit the 'retry' button once it times out which usually happens after about 100-150 emails have been sent. I've actually scripted this action - so now when I send, I hit the send button as well as fire off a shell script that spoofs the 'refresh' request. It takes about 10 minutes to complete the campaign.
I'll echo the other complaints about the quality of the code - it seems to suffer from a copy / paste syndrome as the database connection logic is repeated in every file, but it could also be that the author has obfuscated it as well. It does do a phone home and some other stuff via some base 64 encoded code. I've unwound it in an attempt to fix another issue I thought it was causing (it wasn't). Here it is if you are curious:
I'll give Sendy credit for getting me this far, though I would like to replace it someday with something I feel I have more control over. Not sure the OP or the other suggestions listed here offer exactly what I'm looking for.
It's the year 2016. It's ridiculous to not have tests. Why even run it? You can't change it and be sure you didn't break something/anything without testing in production.
The page you see at mailtrain.org is what you get once you install and run the Mailtrain software from GitHub - it's not a homepage, it is an actual Mailtrain installation that I use to send out my newsletters
I'm still testing it and it looks very solid. After going through the cod for a few hours, I'm pleasantly surprised.
One money saving feature: you can setup SES and Sparkpost, for instance, and configure it to send more emails from Sparkpost than through SES, that way you can save money using free 100k emails/month plan from Sparkpost.
Me too. The Mailwizz guys are really good at coding and structuring things, but their marketing skills are not so good. I only found about it by searching for mail templates on codecanyon.
I believe you should invest in marketing, it's really hard to find about your product in the first place. All your competitors are weak, but they do seem to do better marketing.
It does matter a lot since it brings a solid foundation to build on, and that foundation is well tested, gets regular updates and makes it easy to create a super extensible application.
Most of users that look for such self hosted solution will know that using a framework is better than "pure php" because they get well tested code.
The use of a framework Does not mean the code is better, i would argue the opposite, many a new dev learns on a framework and often doesn't know the underlying language, whereas choosing not to use a framework is usually a conscious decision based on evaluation of actual needs.
Since the product is self hosted and comes with the source code, if you want, you can customize it, fix bugs yourself, write custom features for clients, write extensions to sell, etc.
If it's built on a solid framework, and you study the framework, you'll know where to look when you want to change a model, a controller, so on.
That is true, to a certain extent. However, if you're willing/able to use Amazon SES, you can save a huge amount of money, while having comparable deliverability. Of course, with deliverability, there are no guarantees. Some email services do worse than self-hosted email, some do better.
For most people with smaller lists (say 20k), the cost savings is not that significant. But, if you have a large list, providers such as Mailchimp are prohibitively expensive.
In case of Mailtrain, Sendy, MailWizz et al. you do not send messages directly but through a paid relay, so the blacklisting is not an issue. The price is usually quite low, for example using Amazon SES costs about $100 for million messages. If the volume is low as well, then several providers have generous free tiers available (SparkPost offers 100k free messages/month), so it might not cost anything at all. SaaS providers offer convenience – you do not have to install, maintain or upgrade anything, "it just works" but in the end most of these use also a relay provider like Amazon SES to do the actual delivery
I think the same of you. Mail marketing is really difficult. That's why we don't have lots of competitors, and this is why (I believe) Mailchimp is so expensive.
Self-hosted, works with Amazon SES. I have many clients that send many thousands of messages using the cheapest of Bluehost shared hosting accounts, when you also use Amazon SES as the mail service. (Things like Board of Realtors groups that post realty available to a discussion list)
> Have you ever considered offering it as a SaaS ?
Sure, I've thought about it;
> Why not?
I'm probably at the limit of what I, as the sole developer of the project can do (or want to do) alone, while also performing my other tasks - such as my excellent product support for my clients.
To move to also running a SaaS, I would need a highly motivated partner to help with the workload of development some of the additional parts of the business.
For now, I'm pretty happy with existing in the niche I've made for myself: mostly those individuals and small businesses who find a SaaS like MailChimp too expensive to use. Really - all you need for something like Dada Mail to get started is a shared hosting account - something "Joe's Italian Restaurant" down the street probably already has. If you want to move up, hooking up Amazon SES is a no brainer and solves most all of the problems with sending out messages to a mailing list.
I also enjoy the quality of life given to me knowing I don't have to make sure the service is running 24/7. I live simply, and there's many projects and goals in my life that do not directly involve software.
You should toss your project on Baqqer, open up some attention and possibly get some pledges to help fund you finding more resources to grow your project.
Good to see someone solving this! I used fiesta.cc with a couple groups years ago, and then it shut down. I've thought about building a clone, but could never prove a reliable customer base. I look forward to trying it out!
I do get asked this question a lot, and I don't actually know what people are asking for. Perhaps a better question would be, "Does it have a wordpress plugin that allows me to do $some_intesting_thing"
If you have a Wordpress site (or any site design), Dada Mail has a templating system to allow it to take on that design,
I am not a Wordpress plugin developer. There is a proof of concept API to some often-used functionality, as well as clients written in a few different languages to show you have to use the API. A wordpress plugin developer would need to talk to me to see what else they would need, and build upon what's already available. I'm listening!
FYI: You can always check someone's profile and see if they have info listed, more efficient and cuts down on extraneous comments and waiting on replies.
I was looking for something like this. The perl code is quite easy to follow. Nice project!
I will try and install it on webfaction and provide feedback for you to add to your docs. :)
Another alternative: SendGrid recently released a much less expensive product [1] which shares most features with MailChimp. Contact storage costs $20 per month for a typical list of 20,000 subscribers (for comparison, storing the same number of contacts at MailChimp would be $150/mo).
We handle deliverability issues ourselves, so you won't have to maintain the reputation of a self-hosted solution. If you have any questions, feel free to reach out to me - my email's in my profile.
Mailgun restricts features unless you pay more money, which increases as your plan increases. I prefer SendGrid's model. They're just different pricing models.
I love SendGrid, Mailgun, etc. but they are not replacements for marketing-friendly ESPs. Marketing ESPs need easy UIs for email composition, segmenting, workflows, database modifications, reporting, etc.
This is really cool! If anyone is interested in easily making a CRM tool to do similar things around list management and bulk mailing (and want control on your environment/stack), I wrote a blogpost on it: http://blog.varunarora.com/how-i-made-a-crm-in-3-hours/
HN, thanks for this thread. I was looking for a self-hosted email newsletter solution for ages, but couldn't really find good ones. The often claimed value proposition of deliverabiliy of all the big names is a myth IMHO, at least in Germany. All the big brand names ALWAYS get put into my spam folder (GMX). Deliverability is more than Gmail lets it pass.
I used a Wordpress Plugin called MyMail for a while, but updating Wordpress is annoying and I'm always afraid that the plugin changes in some way that it breaks.
I will check out the suggested solutions here, but if anyone knows from the top of his head which satisfies these requirements, I'd really appreciate it:
- self-hosted (obviously)
- must be able to use any SMTP server
- preferrably bounce handling via IMAP
- double opt-in, which also can be disabled
- basic API support (subscribe, unsubscribe)
- one click unsubscribe
- Autoresponder capabilities (nice would be: lets me set the specific time of delivery. Say: +1 day after signup at 11am)
- reasonably fast UI
Not required:
- multi-tenant
- template builder
- fancy templates (I will use a very basic email layout)
- GeoIP
- CRM capabilities
I tried Mautic (https://www.mautic.org/) which looked promising, but the UI is so horribly slow, it annoyed me. I don't want to wait 5 seconds for every HTTP request to complete. Especially if I need to set up my email campaigns
Well, deliverability and being self-hosted might not go well together these days. There is a LOT you need to go in terms of setting up the relevant infrastructure; the newsletter software is just a tiny fragment of it. And even if your carefully crafted infrastructure works now, it can suddenly stop, just because some big e-mail hosting service decides they implement a major change (see e.g.: http://tanguy.ortolo.eu/blog/article109/google-ipv6-smtp-res...).
See, I get your point, but my experience is different. All the transactional emails of my SaaS are sent out with PHP's mail() which is obviously the worst infrastructure setup you can get and only a fraction is marked as spam for my receivers.
Same is true for my contact form. A copy is sent to the receiver. This works with my providers SMTP server and again, not that many problems with deliverability.
In my limited experience, sending newsletters doesn't have to be a science project.
Yeah, but you aren't pumping out 250,000 emails from that domain every day, are you?
That's where the whole deliver ability lark comes in, and is why we offload non-transactional emails to the likes of bronto and adestra. One less thing to worry about.
Is there a, say, ballpark figure on the volume you need to send before switching to a SaaS becomes an interesting option? E.g., i have a newsletter with about 300 subscribers that i'd like to host somewhere else, and i don't think will ever reach more than maybe a couple of thousand readers.
I never thought about that. I only see that hosted newsletter software is too expensive. My requirements:
- Send maybe 2,000 transactional emails per week
- Send a few hundred autoresponders ("newsletters") per week
- Once a year or so, send 70,000 emails to all users who signed up
I neither want to have an email list with 70,000 people in it where only a fraction regularly receives an email. Nor do I want to pay several hundred bucks for the occasional broadcast to the entire user base.
And also: I won't use non-German newsletter providers for privacy reasons. This is probably way too strict and kind of an illusion, but that's just how I run my business.
Mailwizz has all the features you want, plus is multi-tenant, has extensions support and can be hosted for cheap. I'm running it on Webfaction for $10/m.
I am the phpList community manager. Anyone used phpList before? We're Open Source (AGPL specifically).
We offer phpList as a hosted service on phpList.com. It's the same code (AGPL) but we deal with deliverability, updates etc. Our clients range from 300 mails a month (free) to high volume senders on VPS with millions of subscribers and mails :)
Yes I did, as self-hosted and for very low volume. Thank you for being part of this project.
However, I stopped using it because of user-experience issues. Interface was not clear and efficient enough, especially when working with templates. Last point, design seemed a bit outdated.
There are UI design changes afoot, I believe. The big changes for self-hosted users recently are the new manual at www.phplist.org/manual and the new forums, discuss.phplist.org (using discourse, which I adore!)
I think for small volume it's always better value to use the .com service. Using self-hosted is a lot of work for small volumes. I always used .com for my small business (years before I worked for phpList too) :)
Does anyone know of an open source Mandrill clone, preferably self-hosted? Finding someone to do newsletters or simply doing via traditional mailing-list software is pretty easy, but finding an alternative for Mandrill is a bit harder.
We have maybe ten different email templates, in 7 different language. Mandrill makes that sort of easy, but their reason decision to require a Mailchimp account (which we don't need) has made us look for alternatives. We even considered building something in-house, but it seems like something someone else would already have done.
We looked down this path, but ultimately just didn't want to be responsible for SMTP servers.
Now we're with Sparkpost, the rate was actually a touch cheaper (by maybe a few hundred a month on a several thousand dollar bill). There are plenty of features they don't have (or don't have working correctly when we tried):
* Css inlining, we opted for doing this ourselves after trying theirs.
* Account sending limits. With mandrill this was managed based on our usage, with Sparkpost, it needs to be manually adjusted through support (hint: you also need to monitor this, they won't warn you if you near limits).
* Tagging of emails is different, less searchable but generally its better in some ways and worse in others.
We're building a self-hosted service similar to Mandrill for bulk and transactional email on https://highrisehq.com. Mandrill was the only service that offered unlimited sub-accounts each with their own reputation and quota. We relied on that to offer bulk email to our customers, and since there’s no good alternative we need to build it in-house. In addition to sub-accounts & quota management, we needed open+click analytics, templates, and a fast RPC api. We also don’t want an email provider going dark on us again, so it’ll support sending from a pool of email providers. Mailgun is the one we’re using first (it’s been great so far) then likely SES.
We’re going to open source it at some point. It’s written in Go with a Postgres backend so deployment is straight forward — should be able to get decent mileage off a single Heroku dyno. I look forward to sharing more soon!
93 comments
[ 2.7 ms ] story [ 164 ms ] threadFunctionality wise, it seems the same.
ninja edit: it made me sick because it was ugly, didn't feel secure, and it took me ages to make the changes I needed to make.
It has never sent to completion on its own, it always requires me to monitor the queue so I can hit the 'retry' button once it times out which usually happens after about 100-150 emails have been sent. I've actually scripted this action - so now when I send, I hit the send button as well as fire off a shell script that spoofs the 'refresh' request. It takes about 10 minutes to complete the campaign.
I'll echo the other complaints about the quality of the code - it seems to suffer from a copy / paste syndrome as the database connection logic is repeated in every file, but it could also be that the author has obfuscated it as well. It does do a phone home and some other stuff via some base 64 encoded code. I've unwound it in an attempt to fix another issue I thought it was causing (it wasn't). Here it is if you are curious:
https://gist.github.com/fowkswe/20a7c4c03835de183ff68fe5ab1a...
I'll give Sendy credit for getting me this far, though I would like to replace it someday with something I feel I have more control over. Not sure the OP or the other suggestions listed here offer exactly what I'm looking for.
> Alpha-grade software. Might or might not work as expected. Awful code base, needs refactoring. No tests. No documentation.
Fills a need. Everything else sucks.
It's the year 2016. It's ridiculous to not have tests. Why even run it? You can't change it and be sure you didn't break something/anything without testing in production.
> No documentation.
The 1990s called. They want their code back.
I use sendy.co, but its WYSIWUG editor is not great compared to MailChimp -- that is the only real thing I think MailChimp has over sendy.co.
Main differences: built using a PHP framework, has extensions support, works on SES and competitors and is cheaper.
Thank you!
One money saving feature: you can setup SES and Sparkpost, for instance, and configure it to send more emails from Sparkpost than through SES, that way you can save money using free 100k emails/month plan from Sparkpost.
It is multi-tenant and you can have sub-accounts with their quotas.
You can even use it to roll your own SaaS with the included user management and billing features.
I believe you should invest in marketing, it's really hard to find about your product in the first place. All your competitors are weak, but they do seem to do better marketing.
[1] http://www.yiiframework.com/
Most of users that look for such self hosted solution will know that using a framework is better than "pure php" because they get well tested code.
If it's built on a solid framework, and you study the framework, you'll know where to look when you want to change a model, a controller, so on.
For most people with smaller lists (say 20k), the cost savings is not that significant. But, if you have a large list, providers such as Mailchimp are prohibitively expensive.
Self-hosted, works with Amazon SES. I have many clients that send many thousands of messages using the cheapest of Bluehost shared hosting accounts, when you also use Amazon SES as the mail service. (Things like Board of Realtors groups that post realty available to a discussion list)
I have known your project for a very long time now.
Have you ever considered offering it as a SaaS ? Why not?
Sure, I've thought about it;
> Why not?
I'm probably at the limit of what I, as the sole developer of the project can do (or want to do) alone, while also performing my other tasks - such as my excellent product support for my clients.
To move to also running a SaaS, I would need a highly motivated partner to help with the workload of development some of the additional parts of the business.
For now, I'm pretty happy with existing in the niche I've made for myself: mostly those individuals and small businesses who find a SaaS like MailChimp too expensive to use. Really - all you need for something like Dada Mail to get started is a shared hosting account - something "Joe's Italian Restaurant" down the street probably already has. If you want to move up, hooking up Amazon SES is a no brainer and solves most all of the problems with sending out messages to a mailing list.
I also enjoy the quality of life given to me knowing I don't have to make sure the service is running 24/7. I live simply, and there's many projects and goals in my life that do not directly involve software.
Disclosure: I'm a dev on this.
If you have a Wordpress site (or any site design), Dada Mail has a templating system to allow it to take on that design,
https://www.youtube.com/watch?v=9-EEvGVCJl4
and adding something like a subscription form is an easy copy/paste.
https://www.youtube.com/watch?v=4Cc3c4NTO2U
I am not a Wordpress plugin developer. There is a proof of concept API to some often-used functionality, as well as clients written in a few different languages to show you have to use the API. A wordpress plugin developer would need to talk to me to see what else they would need, and build upon what's already available. I'm listening!
http://dadamailproject.com/d/features-restful_web_services.p...
We handle deliverability issues ourselves, so you won't have to maintain the reputation of a self-hosted solution. If you have any questions, feel free to reach out to me - my email's in my profile.
[1] https://sendgrid.com/solutions/email-marketing
Disclosure: I'm a SendGrid engineer.
From the Call-to-actions on the website, to the mass mailing and statistics on emails. (with a good email template composer)
I am at a point now with Ghostnote alone that my mailing list on mailchimp is costing me $150 a month.
The project is healthy profitable but its still maybe a $1K a year when all comes to all. Money I would rather spend on other things.
I used a Wordpress Plugin called MyMail for a while, but updating Wordpress is annoying and I'm always afraid that the plugin changes in some way that it breaks.
I will check out the suggested solutions here, but if anyone knows from the top of his head which satisfies these requirements, I'd really appreciate it:
Not required: I tried Mautic (https://www.mautic.org/) which looked promising, but the UI is so horribly slow, it annoyed me. I don't want to wait 5 seconds for every HTTP request to complete. Especially if I need to set up my email campaignsSame is true for my contact form. A copy is sent to the receiver. This works with my providers SMTP server and again, not that many problems with deliverability.
In my limited experience, sending newsletters doesn't have to be a science project.
That's where the whole deliver ability lark comes in, and is why we offload non-transactional emails to the likes of bronto and adestra. One less thing to worry about.
And also: I won't use non-German newsletter providers for privacy reasons. This is probably way too strict and kind of an illusion, but that's just how I run my business.
We offer phpList as a hosted service on phpList.com. It's the same code (AGPL) but we deal with deliverability, updates etc. Our clients range from 300 mails a month (free) to high volume senders on VPS with millions of subscribers and mails :)
I think for small volume it's always better value to use the .com service. Using self-hosted is a lot of work for small volumes. I always used .com for my small business (years before I worked for phpList too) :)
We have maybe ten different email templates, in 7 different language. Mandrill makes that sort of easy, but their reason decision to require a Mailchimp account (which we don't need) has made us look for alternatives. We even considered building something in-house, but it seems like something someone else would already have done.
Now we're with Sparkpost, the rate was actually a touch cheaper (by maybe a few hundred a month on a several thousand dollar bill). There are plenty of features they don't have (or don't have working correctly when we tried):
* Css inlining, we opted for doing this ourselves after trying theirs.
* Account sending limits. With mandrill this was managed based on our usage, with Sparkpost, it needs to be manually adjusted through support (hint: you also need to monitor this, they won't warn you if you near limits).
* Tagging of emails is different, less searchable but generally its better in some ways and worse in others.