Earlier this month I talked about some of the lessons I’ve learned as a three year blogger and one of those points talked about why my email open rate (to subscribers) was only about 30%. I went through a phase where I bribed people to sign up for my list with giveaways. You know, they sign up for my email list in exchange for a giveaway entry.
So, naturally, each time I hosted a giveaway tons of people signed up with dummy email accounts (accounts they never actually check). So they never read my blog post emails and it skews my stats.
Brittany made a great comment in response to this. She said:
Re: your email subscribersβ¦do you ever scrub your list? A few times a year, I clear out subscribers that havenβt opened any emails in the past 90 days. It helps with deliverability and open rate A LOT and has kept me out of the paid tiers for services priced on the number of subscribers.
What a great idea! …And yet, also kind of terrifying.
First I had to code something to find my “inactive” subscribers.
I use my own, custom code email subscription plugin. Since it’s my plugin, I don’t have fancy built-in features already. Everything I want to do, I have to code myself. But that’s also the beauty of it. It means that I’m only limited by my own abilities.
My subscribers are stored and managed in WordPress (via my plugin). My plugin also figures out when an email needs to be sent, who it should be sent to, and what the contents of the email are. But the actual email itself is delivered by Mandrill. Mandrill is also kind enough to add tracking to the emails, to figure out which people open and click inside each email. Plus, you can export this sending data from Mandrill into a CSV file.
So, I exported those tracking statistics from the last 30 days from Mandrill. This gave me a CSV file containing data from all outgoing emails in the last 30 days. Each row had: the email address an email went to, how many times that person opened that specific email, and how many times that person clicked in that specific email. Then, rinse and repeat for each subscriber and each email they received in the last 30 days. The file was 2.63 MB and about 20,000 rows.
This is where the coding came into play.
I had to decipher a CSV file with 20,000 rows!
I had to code a way to decipher this CSV data. Because there’s no way in hell I’m going through it manually and tallying up the open counts for each email address. I coded a script that:
- Allowed me to upload the CSV file.
- Went through each row of the file and added the email to an array, along with the number of times they opened that email.
- If an email already existed in the array, it added the open count to the count that was already stored.
So, in the end, I had an array of email addresses, along with the number of times each person opened my emails in the last 30 days. It was kind of cool because I took this opportunity to figure out who had opened my emails the most. My dad was in the top 5 (hi dad!).
The next step was to modify that array so that it ONLY contained people who had opened zero emails. So I stripped out all emails unless they had the number “0” associated with them (for 0 opens).
And once I had my final array of people who had opened 0 emails, I outputted the list so I could manually review it. I wanted to see who I was removing!
Finally, I wrote another script that went through each email address in my final array and unsubscribed them from my system.
I lost 260 people
Seeing your number of subscribers go down SUCKS. You suddenly feel like you’re less successful. But you know what? It’s just a number. And I love numbers, but you know what I love more? Real numbers.
Is someone REALLY a subscriber if they’re on your list but never open your emails?
Hell fucking no!
I got rid of the people who weren’t reading anything anyway. And if they’re not reading my emails, why should I keep sending them?
My open rate went from 32% to 62%
Those are my before and after stats. All I can say is that it’s nice sending emails to people who actually read them. It was a scary move to delete 260 people from my list, but I now have a higher quality list with a more accurate subscriber number.
I haven’t done this yet because good grief does it scare me. I love my numbers! But … I think I’d like to see higher open rates instead. Working up the courage now.
Yeah numbers are nice, but there’s not much sense in clinging onto them if they’re not real and just some arbitrary, inaccurate number!
I love the idea of focusing on real numbers. I would definitely do this.
Good luck Mandy! π
I read just about every email of yours. Some I click and come to your site and some I just read from the email itself. I find everything very interesting and informative. Plus I think you have a great sense of humour. I really do not want to miss anything that could potentially help me with my blog.
Thank you so much Carrie! That means a lot to me. π I’m so glad you’re enjoying my posts. It’s amazing to see people say that!
Well!
This made me think, and be a bit worried too, not in any concern for my own blog, it’s fairly new and I haven’t really done anything about followers.
So, if you could just clarify for me please. You can tell if I simply read your newsletter straight from email? I read all the ones I’m subscribed to, just like this one, but often I don’t proceed to the blog itself. I’d hate to be deleted!
Can all bloggers tell if emails are opened or is it just your awesomeness with codes and stuff? Anyway, glad you do what you do, it’s nothing I could ever aspire to, I just keep blogging my audiobooks, it’s may way of giving authors a bit of thanks and a bit of promotion. I’m sure you are a great inspiration to loggers a bit more go ahead than me,
So thanks for your work
Bec
Yes, I can tell if you’ve opened my email, even if you don’t click through to the post. With email marketing you can track two different figures:
1. The amount of times someone has OPENED the email.
2. The amount of times someone has CLICKED on a link in the email.
I only deleted people who haven’t OPENED any of my emails in the last 30 days.
Not all email platforms track that information. For example: Feedburner doesn’t. But if you use a more “advanced” service like MailChimp then they will include those tracking codes for you and you can see the results.
I did this a few months ago and it felt so good.
Nice job Jenn! π
That’s a lot of subscribers! Well, I’ve never made that an option in giveaways. Hmm, I did give the option to follow my blog in any way but it wasn’t required and I think of all the email subscribers I have, only 5 at most come from giveaways.
That’s great Bieke! Based on that info, I don’t think you’ll have many bogus subscribers. π I was a bad, bad girl in my early days. Live and learn.
I’m on the email team for a medium sized company with an active email list over 300,000 users. If you don’t open or click in 90 days, you are removed from the email list. I completely agree with everything you’ve said.
That’s an excellent policy, and definitely something necessary with a list that size. I imagine you end up pruning TONS of people every time!
This is a really good idea! If I knew how, I’d totally do it too. It’s so much better to have an accurate number, versus a bigger one. π
Am I correct in thinking you use MailChimp? I know it’s definitely possible to do this in MailChimp. I’m not sure how, but I think they have a knowledgebase you could look through?
I need to do this. I had a ton of people sign up for no apparent reason, I mean, a lot of people from other countries and those same people haven’t been opening them. I need to just delete them off.
You use MailChimp, right? I think they have an automatic way of doing this. π
I only have 19 e-mail subscribers, so no, I haven’t pruned any. *grins* But I’ve also only been blogging for 7 months. I can’t imagine having to go through and not only find all those ghost subscribers, but deleting them. Lots of work even with the coding, and seeing the number of followers drop would have freaked me out too.
Yeah it’s a scary move, but at least now I know I have a quality list. π
Ah! I’ve been wondering about this very thing, so your post was spot on. Thanks for sharing the numbers!
You’re welcome, Katie. π
I don’t have that many email subscribers yet and since I’m still using Jetpack subscription, I don’t necessarily have to remove subscribers for now. I’d really like to keep my numbers real too though so I’ll keep this in mind for when a good clean-out is much needed! Thanks for sharing, Ashley!
Yeah I do pay money for each email I send, but the cost is so negligible. I get 12,000 free emails each month, then after that I pay $0.20 for each THOUSAND emails (so $0.0002 per email).
So I personally didn’t do this for cost reasons, but I totally understand why people on MailChimp WOULD do it for cost reasons. Bogus subscribers could easily bump them up to the next pricing bracket and that’s a huge difference.
Yup, that’s one of the things I love the most since I changed the way I deliver e-mails to subscribers. It’s very easy for me to see who has opened their e-mail, and who hasn’t and it’s really nice to be able to not continue sending them out to people who never open.
I think some of those are actually spam accounts, also – they don’t need to open an e-mail to know that there is new content on my site, and so, they can easily stop by my site to leave long, spammy, stupid comments that akismet catches for me in 99% of the cases π
I actually don’t get many (if any) actual spam accounts since I have my own plugin. Spammers learn how popular signup forms work (like MailChimp) and that’s how they’re able to automate the signup process. But it’s harder for a plugin they have no knowledge of. π
That’s interesting, Ashley π do you also sell your e-mail plugin? *goes to check creative whim page*
Not yet. π I am planning on re-coding it in the coming months and then putting it for sale.
That’s excellent news! I’ll be checking back for that π I’m not 100% happy with the one I’m using, it appears people have trouble unsubscribing.
After you posted this I went in and deleted about 50ish of mine that had either never opened or had not in a while. I may delete more! Thanks for the post.
Great job Mandy!
Ah, the beauty of real numbers, as always. ^^ I think it was a good thing you cleaned house. I don’t have many subscribers, so I don’t need to that just yet, but it’s definitively something to keep in mind. Thanks for the post, Ashley!
But that brings to me a question I still haven’t figured out quite yet: mail subscribers vs RSS. I’ve recently come to load RSS feeds into Thunderbird, where I also manage my e-mails, so now I get some feeds (like yours) twice. It bugs me a little, coz it’s messy. But since I’m a subscriber, I’d like to get both entries in a giveaway. How can you even tell if someone subscribed via RSS (if you don’t use feedburner — I only use the WP integrated option)? Coz I never actually clicked “subscribe” with any of the ones I read, I just added the feed link to my inbox. Do I still count into those statistics, since I often open posts from there? If yes, with what kind of ID? IP address? email address?
Wow, what a scary thought! I’m shivering just thinking of losing that many subscribers, but it really makes sense. They weren’t really true subscribers, so did you really lose anything? Not really. Brave move, and I applaud you for it. I’ve wondered about my unopened rate; maybe I need to try a purge and cleanse.
I did this too! I did it because I went over 1k on Mail Chimp which pushed me into a new payment tier and it seemed ridiculous because so many weren’t being opened, so I deleted about 500 I think to get me under 500 to the much cheaper tier and I haven’t looked back.
I did! We recently switched providers and instead of moving the subscribers ourselves, we asked whomever wanted to keep receiving our emails to re-subscribe. Some did, others didn’t, and we ended up losing 50 of our subscribers. But that’s ok, they obviously didn’t want to keep receiving our updates. π
That’s a great idea Karolina! A lot of people get really scared about moving to new providers and want to ensure that EVERY SINGLE subscriber makes it over. And I get that, but I LOVE what you did! You weeded out all the people who probably didn’t open your emails anyway. No sense in keeping those folks around. π