Ask HN: Review my just launched email newsletter: WeeklyAndroid.com

> We decided not to use a email newsletter service such as MailChimp, because they get too expensive when dealing with a couple of thousand subscribers (as we hope to gain soon :).

Good luck, but the odds are stacked against you. When you get a few thousand subscribers is exactly when you need a service to send the emails for you. You will be blacklisted if you're not extraordinarily careful and even at best your deliverability will still be lower than what you'd have through a professional email service. Spammers have made email hell.

I subscribed, let's hope I get it :). No double opt-in yet...

Thanks!

I totally agree - you need to be extraordinary carefully not to land on blacklists. Nevertheless we think it's doable and give it a try. We try hard to follow best practices and have implemented basics such as using a dedicated IP, SPF entries, reverse DNS and basic policies like only emailing user that explicitly opted in.

Any suggestions for additional precautions are strongly appreciated :)

PS: Thanks for the double opt-in hint, it's on the todo list!

Well you're not double opting in at the moment, which is pretty much the first rule of email marketing. You should be doing DKIM signing as well. And sign up for all the feedback loops you can.

As for the daily grind, you need to pay attention to every single bounced message. They bounce for different reasons and depending on the reason you need to do different things. Make a mistake and you get blacklisted. Even better, the error codes vary depending on the server software so it's not easy to automate.

There's a lot more to it, but as I said I wouldn't attempt to do it myself.

FWIW I saw the Lonely Planet (travel site) online manager speak at an email marketing conference last year, and he said the best thing they had done (in terms of database growth) was go to single opt-in.

Needless to say he got frowns from the 'best practice' people in the room, including the conference sponsors, but I'd say it's a rule not the first rule.

I guess I just don't understand why you would roll your own. What advantages do you hope to gain in the short run when you have no idea if you'll be around for the long run?

Don't let the excitement of the moment overwhelm you. Building your own email sending system (implementing policies to avoid being black listed etc) complete with analytics is a non-trivial endeavor.

9.9 times out of 10 the smart move is to go with a free mailchimp plan THEN try to roll your own when it gets too expensive.

be smart.

You guys are so right - your feedback made me reconsider the current approach and choose to go with MailChimp. Often it's so easy to get intrigued by an idea with interesting code to write ;)

Clickable: http://www.weeklyandroid.com

If you're not going to use a service like MailChimp, I'd highly recommend http://sendgrid.com to ensure email delivery.

They give you an SMTP server and make it real easy. For $10/month you can send 10,000 emails.

However, if you're gonna be serious about an email newsletter, that means you'll eventually want analytics. IMO that's the biggest benefit of a service like MailChimp/MadMimi. Tracking opens and clicks is definitely something you can roll yourself, but I'm not sure if it's something you want to tackle. http://en.wikipedia.org/wiki/Web_beacons

Thanks for the feedback.

For me the real advantages of services like sendgrid and mailchimp are 1) feedback loops with ISP's and spam indexes, 2) analytics and 3) high deliverability. We could roll our own analytics and try on the deliverability, but the feedback loop is totally impossible to get into.

Can you elaborate on the service you did end up using? Congrats on the launch. I'm preparing to start a newsletter service myself with local content at first, see if people bite.

I'm studying a dozen of services and any reason to solidify one against another will help. I'm mostly concerned about filtering/categorizing the inputs into geographical or zip codes and I can't figure out the best service with flexibility like that. Campaign monitor is the closest to that.

Also, are you bootstrapping the initial costs alone? server, email splash etc. and have you registered the business yet? You can probably claim those expenses.

Too many questions, sorry! :P

We are writing our own backend in Python, with the idea of publishing it under an open source license later on. Perhaps there are other open source solutions for that too.

Since I have multiple servers running for various projects there are no additional costs but the time I need to put into each week. The editors are helping lot too.

As for business registration, the newsletter is just one project of my primary (umbrella) startup.

Alright, I'm in. One question: are you going to have tailored emails based on the categories I selected to be notified about, or are you just gathering this info to help you make decisions about what content to include overall?

For the first few issues everyone will get the same email, and we use the categories to make the overall content decisions.

If there is enough content in the future, we have the idea to let the backend compile the final mail based on your preferred topics (or leave out the unwanted ones).

Update: We've setup a list at MailChimp and updated the signup form on the main website (http://www.weeklyandroid.com) to point to the new home of our mailing list :)

Mail Chimp is absolutely worth it IMHO. Spend your time on your content.