Lessons from running the Ruby Weekly newsletter so far
It's been 6 weeks since I launched Ruby Weekly, an e-mail newsletter for Ruby developers. It's gone well and I've learned and observed a few things.
Subscriber numbers. I previously wrote about how I got 1120 subscribers in 2 days and the subscription rate has tailed off since then. It's now up to 2173 subscribers and I'm happy with that. My short term goal is 3000 subscribers at which level it'll make sense to explore advertising options.
Between August 27 and October 9, 2010, the subscriber count has gone from 1120 to 2173, a growth of 1053 subscribers. In the same period Google Analytics reports 1910 visitors to the site. Even though subscribing is the only thing you can do on the site, this strikes me as a high conversion rate (55.1%).
Kudos. I've had several e-mails and seen quite a few tweets with compliments about the newsletter, its format, and its content. This sort of recognition doesn't come with all of my projects so it's delightful to see and I appreciate it a lot. Knowing that people are "invested" in the project encourages me to maintain a high quality.
Open rate. The "open rate" of a newsletter is the percentage of people who are known to have opened it. MailChimp detects this through an invisible graphic added into the HTML version of your e-mail but, I believe, also through clicks on links where people have images turned off (and if you have link tracking turned on). This has averaged 66%.
Clicks. Ruby Weekly issues typically have about 8-10 links and on a mailing to ~2000 people, I've seen between 2000-2700 outgoing clicks in all. This isn't mindblowing but is far above average compared to MailChimp's stats. I suspect if I were reading Ruby Weekly I'd only click on a couple of the links so I'm reasonably happy with this.
The "weakest link" so far was to Ruby and Rails, an event in Amsterdam is some of the best apartments Amsterdam. I've had poor luck with linking to events on Ruby Inside so this didn't surprise me. From 2000 subscribers it got only 54 clicks overall. This isn't as gloomy as it could be, though. Those 54 visitors are obviously interested in the event and if Ruby and Rails were paying $100 to be in that mailing, say, then even picking up 4 or 5 registrations would be a good return.
Location. Subscribers from the United States by far make up the lion's share of the subscriber base. Taking a recent mailing, there were 964 US opens, 192 in the UK, 93 in Germany, 85 in France, 85 in Canada, 61 in Australia, 50 in Poland, and it tails off from there. I was surprised to see China outweigh Japan significantly at 41 to 16 but Japan seems to have its own sources for Ruby news.
Content. I've had nothing but compliments on the content of the newsletter which strikes me as odd because it's been so haphazard. For one issue, I was on vacation so I wrote 80% of it before leaving and just added a couple of news items on the day (while working over a horrible GPRS connection). No complaints and click rates remained level.
What has surprised me is that the most successful links are frequently to the "Cool but not Ruby" type sections I add when I find a link that might interest Rubyists but isn't necessarily about Ruby. One example was a news announcement about the Amazon EC2 micro instances. It seems there's an appetite for more general open source development ecosystem news than a "Ruby newsletter" should be providing. I am considering branching out to fill this void.