Lessons from running Perl Weekly for the first 6 weeks
Now, after sending out 6 issues, let me go over some of the things I learned.
Let me also compare that with what Peter Cooper wrote after running the Ruby weekly for 6 weeks.
After announcing the first issue more than 200 people signed up within 3 days. (compared with 1120 for the Ruby weekly) After every issue the number of subscribers jumped again. Usually by about 100 people a week, though in the last 7 days it slowed down. As of today there are more than 700 e-mail subscribers and 170 people following the RSS feed.
During the first 6 weeks there were 3,169 visits to the web site.
The top 10 sources are as follows:
It would be nice to figure out what is behind the (direct) entry. Are those people already subscribed clicking on the link in the newsletter itself? Are those people reading twitter in an external tool and not on twiter.com?
It is about 900 subscribers / 3200 visits or 28% conversion rate. As about 68% are "new visits" this means the conversion rate of individual visitors is 900 / ( 3200 * 0.68 ) = 44%. I know some people follow the newsletter on the site instead of signing up. These are the repeat visitors.
There were several ways I tried to reach the audience.
- Posted on my own blog about the service.
- After every issue I sent out a Tweet
- Posted on my Facebook
- Posted on Google Plus
- Posted in one of the two largest groups on LinkedIn
Other people also helped
- It was sent out via Perl Buzz
- It was mentioned on I-Programmer.info
- It was mentioned on Linux Insight by someone else using the name perlweekly.
- It got mentioned on Reddit. Even twice
- Pradeep mentioned it in his blog
- Olaf Alders mentioned in his MetaCPAN roundup
- Some people tweeted on their own
Maybe I missed a few others. Please send me the links!
I got a number of positive comments on Twitter. Several people have tweeted about the Perl Weekly or re-tweeted my announcements. I got a couple of negative comments on how I asked for re-tweets which lead me to a little research on how to get more re-tweets. I'll post about it later.
A few people "liked" my postings on Facebook. More people pressed +1 on Google Plus when I was posting about the new issues there. Some people even shared the link.
I got positive comments in e-mail and even on Reddit!
There was one question raised by several people whether it is not redundant with Perl Buzz. Andy Lester also wanted to know what was missing from Perl Buzz that made me start the Perl Weekly. This led to a short discussion. Andy has also been one of the biggest aids in getting more attention to the Perl Weekly. I hope I can return the favor, as I think what he has been doing is great.
In any case if you need advice on how Land the Tech Job You Love I'll recommend his book.
Let me explain here again what I said to several people already. There are several distinctions between Perl Buzz and the Perl Weekly.
Andy sends out tweets about interesting articles via the @perlbuzz twitter account. He has some 1500 followers there. That's the primary distribution channel. He also posts the collected links on the Perl buzz website. Occasionally he also publishes his own writing on the website.
The Perl Weekly is primarily a mailing list for people who prefer that format, though it also allows people to use an RSS reader. It also includes a few lines of comments for most of the links.
There might also be differences in the things Andy and I think important to include .
Finally, I think the best answer is from Andy: "Redundancy is not necessarily a bad thing."
This sounds way too business-like and marketingish and we don't want to go there. I collected some of the comments I got on Twitter and other channels and put them on the front page. They show what the readers say about the Perl Weekly.
If you want to be there just send a nice tweet about the Perl Weekly and I can include you too :).
People asked for a text version of the newsletter so now every message includes both the HTML version and the text version.
Open Rate - Click Rate
Peter had that information but as I am still using a simple Mailman mailing list I don't have such data. Don't worry though I am working on my own system to distribute the messages that will also allow me to log which links were clicked and I'll post about that later.
Compared with Ruby Weekly
The Perl weekly got 705 subscribers in 6 weeks (+ 160 RSS subscribers) while the Ruby Weekly got 2173 in the same period of time. Why the huge difference?
There can be several explanations:
For one, when Peter Cooper launched the Ruby Weekly he was running two very successful Ruby related sites Ruby Inside and RubyFlow. They both have some 20-30,000 subscribers. He posted about the Ruby Weekly there. Compared to that my own blog, where you are reading this, attracts only a few hundred readers a day.
On the other hand if I understood correctly Mark Keating said at YAPC::EU that the Ironman has 4,000 visitors a month. That's still far away from the 700 the Perl Weekly has but compared to the 25K subscribers of Ruby Inside that itself might be an indication that there are lot fewer people interested reading about Perl than about Ruby.
It could be that the things I select for the Perl Weekly are not that attractive to many people, though that would probably also lead to people unsubscribing. That has not happened yet.
It might be that the web site is not convincing enough to get a higher percentage of people to sign up. I made many changes since the launch but there are not enough visitors to run A/B test that could give significant results.
Finally it might be that the way I was trying to promote the Perl Weekly was not good and I have not reached many potential readers. Yet.
ConclusionIn any case this was an interesting experience so far, but it was only the beginning. The real trick is being persistent doing this. So expect the next issue on Monday.
Just make sure you subscribe to the Perl Weekly.
If you have any comments regarding any of the above or about the Perl Weekly in general, please don't hesitate to share them. I even enabled comments without login, as it seem that was the biggest thing people complained about earlier when I wrote about 3 simple ways to help the Perl community
Let's see how this works out.