For most reasonable applications that is basically a great advertisement for mailchimp, Eg if you rolled your own solution you’d only save eur 500 a year before accounting for the build and maintenance of your solution.
Given the recent revelations that MailChimp is fact checking (reading) your email content and deciding whether they should not send it (i.e. the contemporary corporate censorship cause)... there may be more than cost for self hosting business applications
And then, it is not even your own solution, but AWS SES.
I have been running mail servers for 20+ years, sending volumes that would have made me multiple times bancrupt if I was doing it via Mailchimp or SES.
At a previous job, all API errors were emailed to internal email lists.
The dev monitoring the API health kept track of system health by looking at the number of unread messages in his inbox. If the number of unread messages went up by < 100,000 in a day, the API was doing fine. If there were a couple million new messages, it means we were having problems.
So... I can definitely see how this is plausible, if not uh, "best practice".
An engineer should recognize what is worth their time. If setting this up takes more than 5hour a year of time to setup and maintain then it’s not worth it. You’ll be wasting money
Not eveyone earns 100€ an hour, especially in Europe. The newsletter has a French name ("Le courrier du Hacker"), and in France 45/50k€ a year is a good salary. That's around 14/15€ an hour. Which means ~33 hours a year of maintenance.
There are many ways to measure the cost of something like this, but a blind look at average salaries is probably the worst. It doesn't factor in overhead or opportunity cost. In most companies, having a developer spend a week on maintenance of a system which isn't a core business value doesn't make sense. They would be much better utilized on value added work. I guarantee you most engineers could delivery way more than $500 in savings in a week.
As I said here https://news.ycombinator.com/item?id=27259123 this isn't a company, it's a newsletter so more of something like a side project. I don't think you could deliver more than 500€ of savings in a week because the email costs were the main costs. I agree with you than within a company this would be different, but this isn't a company.
My comment was targeted to the person I replied to, which I interpreted as "managers only think in terms of time saved where as developers if merely left alone would roll it their own and save time and money", which is ridiculous. An engineer should think like their manager, and the manager should think like the engineer. A cost/benefit should be realized, and I think for the high majority of companies, you should not be rolling your own to save 500/year.
A newsletter is so different because merely this blogpost could have generated enough money/attention to warrant doing it, but there's an endless amount of reasoning that can be done to warrant both choices.
But this isn't a company, it's a newletter the author is running. It's important to consider the context of things. Here the only cost is opportunity cost, which I wouldn't know how to evaluate. On the other hand, they save 500€ a year, learned new skills and created content for their blog.
I think it is actually closer to ~16 hours a year assuming the lower end of the range you suggested, and dramatically less if this person makes more than that.
Most simply, I think the typical calculation is if there were 40 working hours per week (8 hours, Mon - Fri) then there are 2080 working hours per year, which is ~21.63 Euro (per hour, assuming 45k EUR) just in salary - other commenters pointed out an employer would have other costs, but lets try and keep it simple as was suggested.
But beyond simple, you don't work on public holidays and Europe tends to have 20-25 vacation days per year as well. That probably adds up to 35 days (280 hours) in France. Therefore the number of hours actually 'at work' is only going to be 1800 a year, so 45k EUR comes to 25EUR per hour.
I imagine this might work on AWS Free Tier but while SES supports ~62k emails per month as 'Always Free', any EC2 instance or network bandwidth is probably billable here. There's probably up to 100EUR a year in cost associated here?
So you have 400EUR of money Mailchimp charged for the service and will never be "apples to apples" because now you're responsible for initial setup and ongoing maintenance. Definitely seems simple in the blog post, but unmentioned was how you backup all the configuration and state, how do you monitor it, most of the features of Mailtrain (e.g. Click Stats) rely on the NodeJS application being operational so is this actually a load balancer, two instances, and the extra complexity of database replication, et al? It is also now on you to make sure it stays patched, isn't compromised, etc.
I imagine there are different deliverability promises with Mailchimp vs. Amazon SES and you might now have to do more managing of that through AWS interfaces?
All of the above needs to fit into ~16 hours a year if we're going to draw this comparison against what you'd make as a salaried engineer. All of the above is time that could have been spent on extra content for the newsletter, or improved content for it, other than this post obviously, or on audience building, on other professional development, or just about anything else.
I think the ROI of writing the article and making everyone debate this on the internet was pretty high, but the title feels like clickbait. This is a straight trade for saving a little money vs. a semi-significant personal time investment.
Here are my calculations: a salary of 45k euros in France means you take home 2618€ a month (according to https://www.salaire-brut-en-net.fr/). A month is ~4.33 weeks, a week is 35 hours. 2618 / (35 * 4.33) gives me ~17.40, so I made a mistake with my initial calculation. Still, it's way less than your 25 euros per hour. Maybe you underestimate the difference between "brut" and "net" and then taxes, which are both pretty high in France.
500€ at 17.40€ an hour is 28.7 hours, 22.9 if we're taking your 400€ (I'll believe you on that, I don't know much about AWS pricing). I think that may be reasonable to take care of it. There's also a big problem: you can't directly transform hours of your life into money. If the author is not paid by the hour because he's a cadre, he had to work 218 days in the year, and that's it, working more hours won't make him directly earn more (more info here https://www.service-public.fr/particuliers/vosdroits/F19261?...). Finding freelance work to expand those 10 to 30 hours (to be very large) isn't going to be easy too. As an aside, if we take 218 days for our calculations, that comes up to ~20€ an hour if they work 8 hours days, and ~23€ an hour if they work 7 hours days.
I do agree that database replication seems like a big issue, and if that's the case maybe switching to RDS could be a good idea.
Anyway, I think the article is relatively good at demonstrating the two extremes, Mailchimp vs everything by hand. Everyone is free to take of that what they find interesting, even if it's "I'd rather pay for Mailchimp".
> I think the ROI of writing the article and making everyone debate this on the internet was pretty high, but the title feels like clickbait. This is a straight trade for saving a little money vs. a semi-significant personal time investment.
I do think that it should be "How I" instead of "How to", but on the other hand "save up x/year" already informs you that the author is trading their time for money. I don't think it's clickbait just because it wouldn't apply to your or other peoples situations. Especially if 500€ is a little money for you, you're not really the target of the article.
Depends who is paying. If it’s your own project then 5 hours of your own time might not be worth that much. If someone else is paying you then both options cost money and the equation changes.
Yes and no. There is some value in the experience gained. That being said, there might be better experiences to be gained (i.e., stick with Mailchimp and learn some other tool).
If you're in the EU you're also saving the GDPR fines. And the author probably saved 500 euros a year on a basic plan. But if you're on the top end plan you would save tens of thousands.
Wasn't that just the latest iteration of "EU pretends no other country in the world has similar national security laws to its own member states", under which multiple EU-US agreements on data sharing have already been struck down and technically any EU business probably can't use any US-based service ever if that involves sharing of personal data? Did this one get to the end of the process and result in a real fine or other penalty?
> Did this one get to the end of the process and result in a real fine or other penalty?
It did end with a judgement that it was in breach but the court didn't fine the company because they barely used Mailchimp and they stopped using it before it got to trial.
GDPR as far as I know applies to anyone collecting data within the EU. But seriously the people who would think this is a great ad for mailchimp because the savings are so low are companies.
Assuming that the newsletter application is a fairly stable release and it's on a package manager or auto updates via docker. The management of that is probably pretty low.
Except mailchimp has dumbed it’s interface down so much that maintaining it takes far more time than installing a package like this or paying a small fee for a similar SES backed service like Mail Octopus.
We also pay around 200USD/mo and are a small local business with some online presence for e-tutorials etc. They were cheap for a while but if you get over 15k contacts it's a crazy price if you are not a heavily monetised business and mostly just want to send newsletters not conversion funnel jank.
How many of these €50/month does it take to reconsider? €50 for hosting, €50 for email, €50 for tracing, €50 for sentiment analysis, €50 for this, €50 for that. By the time you realise, you are burning €6k a year on stuff that takes hours to roll out yourself.
You are suggesting rolling out all of those things yourself will take only a one-time investment of "hours" (meaning fewer than 10?), with no ongoing maintenance?
I think at minimum "hours" per month seems more realistic, no?
I've been running Sendy for a while on my own host. I spend about 30 minutes a year on maintenance time to occasionally upgrade versions. Each upgrade takes about 10 minutes and I tend do it every few months.
Shoutout to postal [1] , for outbound email delivery for apps and smaller well maintained mailing, the likes of SES/Sendgrid can be also be removed from the stack.
While email delivery is complex, and reputation matters, for smaller apps and well established mailing lists, self setup works perfectly fine.
500€ is 5 hours of labor. If you are skilled and everything works right away you may only spend 2 hours on the self-maintained solution which means you save 3 hours or in other words … nothing. You also have to maintain/update the system, maybe spend time teaching other people. It’s not worth the effort.
If you don’t have expert knowledge in the required tools/services, it will take much longer than 5 hours to set up the self-maintained solution. If you are unlucky “you may need the help of a professional sysadmin.” … Again, saving nothing.
It’s not possible to catch an even number of leprechauns. I mean, basic fact checking is what I expect from a HN commenter. This place is turning into Reddit.
Well, if you don't have anything better to do, got the time and have a DIY attitude, then it's a great investment. Also, there are more years to come so Y+1 it still saves you 500.
This doesn't sound like a good deal for the average business with zero experiencing managing their mail.
One spam block and you're done, you'll spend at least a week dealing with that.
Indeed true, if you are on one of those big providers and they're failing you there's not much to lose aside from your own time/money by DIY'ing it. The ESP business is basically all about volume and if you need special handling you're going to get left behind.
I know that one of the very hard problems with automating email is not landing in the spam folder.
I always had the impression that using services like Mailchimp make that much less likely than doing it DIY on top of a more commodity transport like SES. I.e. they stay on top of and evolve with best practices over time so you don't have to.
Not in our experience, we've only used SES to send emails and haven't had any reports of Customers not receiving emails or any of our emails having ended up in their spam folder, although we've never sent any marketing or promotional emails so our flagged rates should be really low.
I moved my mailing list (all explicit subscribes, plaintext content) from Mailchimp to an SES-based solution in part for better handling of plaintext email, which is a total afterthought for most of these tools. Since the switch I have experienced minor hard rejection problems with SES that I didn't before, primarily with Apple Mail and rarely with gmail as well. I wouldn't say it's been a huge problem for me, in part because I don't care that much, but as far as I can tell Apple Mail was rejecting 100% of what I sent for a while while gmail just sporadically randomly rejects an email from time to time, not even the same email across multiple recipients.
We started having this recently when sending quotations to customers. We ended up sending attached PDFs instead,so Apple Mail won't complain anymore. It can be either content or how the domain is setup,still not sure to this day.
> I.e. they stay on top of and evolve with best practices over time so you don't have to.
> Is that in any way accurate?
For the most part, yes.
It's all about delivering as much information as possible to aid the receiver with the spam detection process. The goal is to reduce false positives/negatives. The beauty of this is of course that if you are in fact sending spam, this will work against you.
Mailchimp does this very well, they actually have people working for them with 'email deliverability expert' as their job title.
The problem is that rolling a simple email sending solution is trivial to set up, but hard to get right. It takes surprisingly significant engineering resources to build a solution that complies with all the relevant RFCs, scales nicely and follows new developments. Resources that are probably much better spend elsewhere.
My job is helping customers optimise their email infrastructure [0], I have analysed hundreds of domains, and my observation is that the vast majority of home-grown email solutions lack even the basic provisions for deliverability.
My advice: don't roll your own email solution, unless you can afford full-time engineers to do so. There certainly is a tipping point (as always) where building your own solution is more cost effective as a SaaS solution, but that is certainly not at 500€/Year.
The problem is that you actually are spamming. Almost no one wants the junk that comes out of mailchimp. It’s just they are big and powerful enough to get their spam in inboxes rather than the spam box where it belongs.
I used to use Pardot ( B2B marketing software) at work, which is miles ahead of MailChimp. While there are certain things you can't do by default (e.g. email to info@domain.com,etc) , the spam filters are still kind of blackbox,so the content can have a huge play in addition to marketing software server's reputation,etc.
>The drawback is having to monitor blocklists and the reputation of your server IP address.
Which sound like big drawbacks for anything reasonably serious. Maybe if you're doing this at scale for clients but, then, you're potentially exposed to clients who can cause problems that affect other clients.
Sadly, this says more about how broken GMail now is than anything else.
Real Soon Now(TM), the tech community needs to start a serious campaign to shame and isolate email providers that do not reliably deliver legitimate emails, regardless of how high profile they are. It probably needs meaningful government/regulator backing to get anywhere. The odd false positive on spam/malware filtering is understandable, it's not a perfect science, but the reality today is that a lot of these services simply don't work properly and the unreliability is probably causing some degree of damage to huge numbers of people and businesses. Email is far too important for far too many things to let the reliability continue to degrade according to the arbitrary whims of a few large mail services and to permit oligopoly effects to lock new or small-time providers out of the system almost entirely.
But the users might make different choices if they knew how often mail they probably want to receive isn't reaching them. The problem at the moment is that the failures are largely silent from the user's point of view. To be fair to GMail, although it's obnoxiously aggressive with classifying things as spam, as far as I know it doesn't silently drop incoming mail without notifying either the sender or recipient, which is the worst thing that some major ISPs do.
I have had customers contact my business to ask for help with something, and we send them a quick reply answering their question, and then the next day we get a frustrated follow-up asking why we are ignoring them, and this continues for another round or two until they give up on buying something from us or cancel a subscription.
I have had friends send me a reply a month after a big party saying they're sorry they didn't make it but they only just found the invitation and the follow-up I sent them in their junk folder.
Those sorts of cases are annoying and seem to happen all the time these days, but imagine if a missing email was something really important from your doctor or your kid's school.
And you don't think it's a problem that many people might not receive some legitimate emails at all? Or you just don't think that anything can be done to improve the situation now?
But again, the point is that the users often don't know how badly their email is broken.
Unless the user is actively expecting to receive an important mail around a certain time and therefore likely to notice its absence, the failures are either quiet (messages dropped in junk automatically that may or may not be noticed by the recipient later) or totally silent (as with ISPs that drop mail they deem inappropriate entirely, without notifying either party).
Even if someone is expecting a mail, as the experience with business customers I mentioned demonstrates annoyingly often, the failure to receive that email may be assumed to be the fault of the expected sender, not the recipient's own mail service.
I don't really buy that excuse any more. This is a business that wants to have heavy, fast-moving objects automatically navigating complex and unpredictable transport networks using its software. For that matter, this is a business whose golden goose is prioritising content people most want to see on the scale of the entire WWW, which has no shortage of junk sites that exist primarily to game the system. I'm reasonably sure that same business can be smarter about spam recognition than defaulting to assuming new mail from a source it hasn't seen before is bad, which seems close to the practical output of its existing system. But as with so many of the current dominant tech businesses, until someone poses a credible threat to its existing business model and revenue streams, it has no reason to invest in fixing problems even if it could obviously do better.
> To send 4k emails a month a 1€/month VPS running Exim is enough and takes 1h to set up.
You can use a VPS for web projects. For e-mail you'd want a dedicated server with a separate IP that is going to cost you €40/month. Yes it may take 1h to set up, but gaining reputation takes time - you definitely won't be able to reach all recipients within one hour of purchasing the server.
No VPS has a shared IP address. The real problem is that the pool may have been dirtied by previous customers. But once you get a clean IP you can keep it for as long as you keep the service running.
I sent emails from a VPS for years without issues. I stopped purely because Fastmail had better software and cheaper storage.
Hopefully as IPv6 takes over, the contamination effect due to some unknown party who might once have used the same IP address will become less of an issue. Although you have to wonder whether the kind of big mail service that will swing the blacklist axe first and probably not ask questions later with IPv4 addresses is just going to start blocking entire service provider ranges with IPv6 to counter that, causing the same problems for legitimate senders with new systems anyway.
Absolutely overkill to have a metal server at that cost to send emails. A cheap VPS (which always comes with its own IP) is fine in most cases.
From experience I can confirm the reputation build-up requirement though. And all that reputation can get ruined by a single blacklisting - even if you manage to get the server de-listed later the damage is done.
If you're ready to pay 5$ each month for each thousand subscribers, Buttondown seems like a great tool. It does offert paid subscriptions as well if you want to earn money with your content.
He has clients with over 100k+ people on their list and has put in a lot of thought and leg work to ensure good deliverability rates. It's also been running for ~5 years.
I don't think there is. NodeMailer kind of gets close by letting you request Delivery Status Notifications, but doesn't have webhooks or handle them for you: https://nodemailer.com/smtp/dsn/
The closest thing I think exists would be Sendy, but it's a closed source paid app.
CiviCRM can handle bounces (fetches them from a mailbox using a VERP), but for spam complaints, you need to own your IP block in order to manage feedback loops, which is another level of complicated.
I've been working on my own startup for a few months now, I'd say email is one of these things where I really don't mind paying for, even though I hate how expensive Mailchimp/Mandrill is.
Speaking from personal experiences on non-dedicated IP plans on both SendGrid and SES, SES is much better in deliverability of transactional emails, esp. when sent to .edu domains.
I felt like SES is more careful in vetting their customers to prevent shared IPs being polluted.
I feel administrators for EDUs (for example) know how to easily block all of sendgrid or mailchimp but don’t know how to block all of SES without blocking the Dean’s Amazon receipts.
If you don't want to pay MailChimp prices but don't want to make the full leap to self-hosting, EmailOctopus Connect is a nice middle ground.[0]
It's similar to Mailtrain in that it uses your Amazon SES account to send the mail, but it's a managed SaaS so you don't have to worry about keeping your server patched or accidentally wiping out your subscriber list.
I've been migrating all my lists out of MailChimp and into EmailOctopus over the past six months, and I've been happy with the service. One nice perk is that you can disable view and click tracking, giving back some privacy to your subscribers. Their API is also the nicest of any providers I looked at.
Can you elaborate on why SES is shit? I've been using it for about 6 months now to run a small newsletter and haven't experienced any issues with deliverability.
But if you do know what you're doing - like myself and many other engineers - you don't need hand holding.
I would not consider SES "shit" because it does not baby-sit you throughout the setup process. The documentation was solid, and I had a lot of freedom to customize a solution specific to my use case. Overall working SES has been an enjoyable experience.
But maybe my mind will change 1 billion emails later haha we'll see.
I agree. When testing the various SES was more involved but seemed to work fine. If you follow the standard procedures and aren’t a spammer (even a respectable spammer) it should work fine.
We sent few million emails using SES over few weeks, Sendgrid's deliverability was around 12 times better than that of SES.
Don't get me wrong, it might be perfectly fine choice for smaller projects but it is not a solution you would rely for scale. You have to build your own mailservers with warmed up IPs.
Can anyone with experience compare Mailtrain / Mautic / others?
I did a quick Mautic setup on a DO droplet a while back. It felt a little like beta software, a lot of careful steps to get things working, some things didn't work, updates and some actions on a clean install gave me PHP errors, Amazon SES API wasn't built yet so sent super slow via SMTP only, etc. And it leaned way harder into marketing automation as I should've expected from the name (JS on page to track users then associate them if/when they sign up later). This was at least a year ago and they've had a major release since so might have fixed some of these but I'm still trying to save money for a small business who can't afford $200/mo for sporadic newsletters and welcome series with free stuff.
Clearly OP's newsletter is a labor of love without much earning potential (or not much effort in that regard), otherwise delegating email delivery to a service like MailChimp is the wisest move.
If you're a hacker trying to launch a business, building your own email service to save $50/month is a ridiculous distraction from getting to profitability.
Agreed. I spend over $1,000/mo on various software for my business. Occasionally I'll trim the fat, but saving a few bucks here and there isn't worth it when energy spent elsewhere makes far more.
I wrote listmonk[1] after running into difficulties scaling existing open source list managers including Mailtrain to millions of subscribers. It is a single binary app that can run on a potato VM instance.
For folks looking for a Rails engine that doesn’t ship with a GUI, I built Caffeinate to manage/create/send scheduled email sequences/drip campaigns: https://caffeinate.email
It could also be used for creating and managing a newsletter, but doing so is undocumented as of writing.
The problem I find quite often with a lot of these tools is that whilst it's using AWS SES to do the sending it can be a bit cumbersome or sometimes not at all possible to see which of your emails are being delivered, which are bouncing etc. There's some cool options out there like sesmonitor.com and sesdashboard.com which add that extra layer
113 comments
[ 5.4 ms ] story [ 182 ms ] threadWhy would I go though all this for 500 euro a year.
Why doesn't OP deploy this as a service for 15$ a month ?
https://techstartups.com/2020/10/30/forget-big-tech-censorsh...
I have been running mail servers for 20+ years, sending volumes that would have made me multiple times bancrupt if I was doing it via Mailchimp or SES.
Also, 20 years ago running your email server was way easier than now
The dev monitoring the API health kept track of system health by looking at the number of unread messages in his inbox. If the number of unread messages went up by < 100,000 in a day, the API was doing fine. If there were a couple million new messages, it means we were having problems.
So... I can definitely see how this is plausible, if not uh, "best practice".
A newsletter is so different because merely this blogpost could have generated enough money/attention to warrant doing it, but there's an endless amount of reasoning that can be done to warrant both choices.
Most simply, I think the typical calculation is if there were 40 working hours per week (8 hours, Mon - Fri) then there are 2080 working hours per year, which is ~21.63 Euro (per hour, assuming 45k EUR) just in salary - other commenters pointed out an employer would have other costs, but lets try and keep it simple as was suggested.
But beyond simple, you don't work on public holidays and Europe tends to have 20-25 vacation days per year as well. That probably adds up to 35 days (280 hours) in France. Therefore the number of hours actually 'at work' is only going to be 1800 a year, so 45k EUR comes to 25EUR per hour.
I imagine this might work on AWS Free Tier but while SES supports ~62k emails per month as 'Always Free', any EC2 instance or network bandwidth is probably billable here. There's probably up to 100EUR a year in cost associated here?
So you have 400EUR of money Mailchimp charged for the service and will never be "apples to apples" because now you're responsible for initial setup and ongoing maintenance. Definitely seems simple in the blog post, but unmentioned was how you backup all the configuration and state, how do you monitor it, most of the features of Mailtrain (e.g. Click Stats) rely on the NodeJS application being operational so is this actually a load balancer, two instances, and the extra complexity of database replication, et al? It is also now on you to make sure it stays patched, isn't compromised, etc.
I imagine there are different deliverability promises with Mailchimp vs. Amazon SES and you might now have to do more managing of that through AWS interfaces?
All of the above needs to fit into ~16 hours a year if we're going to draw this comparison against what you'd make as a salaried engineer. All of the above is time that could have been spent on extra content for the newsletter, or improved content for it, other than this post obviously, or on audience building, on other professional development, or just about anything else.
I think the ROI of writing the article and making everyone debate this on the internet was pretty high, but the title feels like clickbait. This is a straight trade for saving a little money vs. a semi-significant personal time investment.
500€ at 17.40€ an hour is 28.7 hours, 22.9 if we're taking your 400€ (I'll believe you on that, I don't know much about AWS pricing). I think that may be reasonable to take care of it. There's also a big problem: you can't directly transform hours of your life into money. If the author is not paid by the hour because he's a cadre, he had to work 218 days in the year, and that's it, working more hours won't make him directly earn more (more info here https://www.service-public.fr/particuliers/vosdroits/F19261?...). Finding freelance work to expand those 10 to 30 hours (to be very large) isn't going to be easy too. As an aside, if we take 218 days for our calculations, that comes up to ~20€ an hour if they work 8 hours days, and ~23€ an hour if they work 7 hours days.
I do agree that database replication seems like a big issue, and if that's the case maybe switching to RDS could be a good idea.
Anyway, I think the article is relatively good at demonstrating the two extremes, Mailchimp vs everything by hand. Everyone is free to take of that what they find interesting, even if it's "I'd rather pay for Mailchimp".
> I think the ROI of writing the article and making everyone debate this on the internet was pretty high, but the title feels like clickbait. This is a straight trade for saving a little money vs. a semi-significant personal time investment.
I do think that it should be "How I" instead of "How to", but on the other hand "save up x/year" already informs you that the author is trading their time for money. I don't think it's clickbait just because it wouldn't apply to your or other peoples situations. Especially if 500€ is a little money for you, you're not really the target of the article.
Regardless, it's not simply about time.
It did end with a judgement that it was in breach but the court didn't fine the company because they barely used Mailchimp and they stopped using it before it got to trial.
I think at minimum "hours" per month seems more realistic, no?
I've been running Sendy for a while on my own host. I spend about 30 minutes a year on maintenance time to occasionally upgrade versions. Each upgrade takes about 10 minutes and I tend do it every few months.
While email delivery is complex, and reputation matters, for smaller apps and well established mailing lists, self setup works perfectly fine.
[1] https://github.com/postalhq/postal
If you don’t have expert knowledge in the required tools/services, it will take much longer than 5 hours to set up the self-maintained solution. If you are unlucky “you may need the help of a professional sysadmin.” … Again, saving nothing.
Only if someone will pay you for those 5 hours. Otherwise may as well say they’re worth 37 starfish or 92 leprechauns.
I’m trying as hard as I can.
It's not, see https://news.ycombinator.com/item?id=27259000.
> You also have to maintain/update the system, maybe spend time teaching other people.
I think it's just one person running a newletter, so I don't think teaching other people is a thing here.
And customer service that denies their own problems.
I always had the impression that using services like Mailchimp make that much less likely than doing it DIY on top of a more commodity transport like SES. I.e. they stay on top of and evolve with best practices over time so you don't have to.
Is that in any way accurate?
We started having this recently when sending quotations to customers. We ended up sending attached PDFs instead,so Apple Mail won't complain anymore. It can be either content or how the domain is setup,still not sure to this day.
For the most part, yes.
It's all about delivering as much information as possible to aid the receiver with the spam detection process. The goal is to reduce false positives/negatives. The beauty of this is of course that if you are in fact sending spam, this will work against you. Mailchimp does this very well, they actually have people working for them with 'email deliverability expert' as their job title.
The problem is that rolling a simple email sending solution is trivial to set up, but hard to get right. It takes surprisingly significant engineering resources to build a solution that complies with all the relevant RFCs, scales nicely and follows new developments. Resources that are probably much better spend elsewhere.
My job is helping customers optimise their email infrastructure [0], I have analysed hundreds of domains, and my observation is that the vast majority of home-grown email solutions lack even the basic provisions for deliverability.
My advice: don't roll your own email solution, unless you can afford full-time engineers to do so. There certainly is a tipping point (as always) where building your own solution is more cost effective as a SaaS solution, but that is certainly not at 500€/Year.
[0] Disclaimer: I'm the founder of Mailhardener
The real advantage is to have control over your own infrastructure and ensure the privacy of your reader base.
The drawback is having to monitor blocklists and the reputation of your server IP address.
Which sound like big drawbacks for anything reasonably serious. Maybe if you're doing this at scale for clients but, then, you're potentially exposed to clients who can cause problems that affect other clients.
People who believe this is a viable method to reach the inboxes of thousands of gmail users have never tried.
Real Soon Now(TM), the tech community needs to start a serious campaign to shame and isolate email providers that do not reliably deliver legitimate emails, regardless of how high profile they are. It probably needs meaningful government/regulator backing to get anywhere. The odd false positive on spam/malware filtering is understandable, it's not a perfect science, but the reality today is that a lot of these services simply don't work properly and the unreliability is probably causing some degree of damage to huge numbers of people and businesses. Email is far too important for far too many things to let the reliability continue to degrade according to the arbitrary whims of a few large mail services and to permit oligopoly effects to lock new or small-time providers out of the system almost entirely.
I have had customers contact my business to ask for help with something, and we send them a quick reply answering their question, and then the next day we get a frustrated follow-up asking why we are ignoring them, and this continues for another round or two until they give up on buying something from us or cancel a subscription.
I have had friends send me a reply a month after a big party saying they're sorry they didn't make it but they only just found the invitation and the follow-up I sent them in their junk folder.
Those sorts of cases are annoying and seem to happen all the time these days, but imagine if a missing email was something really important from your doctor or your kid's school.
You're free to try to educate users now, if you like, but most people receive emails only from a small group of hosting companies.
The actual people being negatively affected have chosen an acceptable level of spam false positives.
Unless the user is actively expecting to receive an important mail around a certain time and therefore likely to notice its absence, the failures are either quiet (messages dropped in junk automatically that may or may not be noticed by the recipient later) or totally silent (as with ISPs that drop mail they deem inappropriate entirely, without notifying either party).
Even if someone is expecting a mail, as the experience with business customers I mentioned demonstrates annoyingly often, the failure to receive that email may be assumed to be the fault of the expected sender, not the recipient's own mail service.
You can use a VPS for web projects. For e-mail you'd want a dedicated server with a separate IP that is going to cost you €40/month. Yes it may take 1h to set up, but gaining reputation takes time - you definitely won't be able to reach all recipients within one hour of purchasing the server.
I sent emails from a VPS for years without issues. I stopped purely because Fastmail had better software and cheaper storage.
From experience I can confirm the reputation build-up requirement though. And all that reputation can get ruined by a single blacklisting - even if you manage to get the server de-listed later the damage is done.
https://buttondown.email
I have no affiliation with them but I did speak to the guy who runs Buttondown on my podcast. He went over how it's built and deployed at: https://runninginproduction.com/podcast/70-buttondown-lets-y...
He has clients with over 100k+ people on their list and has put in a lot of thought and leg work to ensure good deliverability rates. It's also been running for ~5 years.
The closest thing I think exists would be Sendy, but it's a closed source paid app.
e.g. email security software driven opens and clicks are registered as human by CC's reporting tools (GUI and API).
One-to-many email solutions are crazy: it's either some crap or super expensive marketing software. We all wish there was something a bit better.
I felt like SES is more careful in vetting their customers to prevent shared IPs being polluted.
It's similar to Mailtrain in that it uses your Amazon SES account to send the mail, but it's a managed SaaS so you don't have to worry about keeping your server patched or accidentally wiping out your subscriber list.
I've been migrating all my lists out of MailChimp and into EmailOctopus over the past six months, and I've been happy with the service. One nice perk is that you can disable view and click tracking, giving back some privacy to your subscribers. Their API is also the nicest of any providers I looked at.
[0] https://emailoctopus.com/pricing-connect
I started using Sendy with SES a few years ago and it was much easier to setup compared to this blog post.
Went from $120 / mo to pennies, + the initial cost of Sendy.
If you're a developer, it's worth it.
Initial price is much higher than Sendy, but Mailchimp recently also added a drip campaigns feature.
[1]: https://mailcoach.app
As someone who has sent billions of emails I can tell you the following.
1. SES is shit. 2. Value of mailchimp is not in its interface. It is in mail deliverability.
Use at least sendgrid.
I would not consider SES "shit" because it does not baby-sit you throughout the setup process. The documentation was solid, and I had a lot of freedom to customize a solution specific to my use case. Overall working SES has been an enjoyable experience.
But maybe my mind will change 1 billion emails later haha we'll see.
Don't get me wrong, it might be perfectly fine choice for smaller projects but it is not a solution you would rely for scale. You have to build your own mailservers with warmed up IPs.
I did a quick Mautic setup on a DO droplet a while back. It felt a little like beta software, a lot of careful steps to get things working, some things didn't work, updates and some actions on a clean install gave me PHP errors, Amazon SES API wasn't built yet so sent super slow via SMTP only, etc. And it leaned way harder into marketing automation as I should've expected from the name (JS on page to track users then associate them if/when they sign up later). This was at least a year ago and they've had a major release since so might have fixed some of these but I'm still trying to save money for a small business who can't afford $200/mo for sporadic newsletters and welcome series with free stuff.
If you're a hacker trying to launch a business, building your own email service to save $50/month is a ridiculous distraction from getting to profitability.
[1] https://listmonk.app
It could also be used for creating and managing a newsletter, but doing so is undocumented as of writing.