a few months ago I noticed, that my Newsletter-openrate is going down. So now I made a quick research and found a lot of articles about list-cleaning.
I would also like to do this, but I am very unsure. Because I have a whole series of people who apparently have 0 opens, but the number is absurdly high and I can’t imagine that. That’s why I wanted to ask about best practices here. How do you do it?
My idea was to look for free users who have had 0% opens in the last 6 months. Would that be reasonable? Because, as I read, the data in the backend may also be incorrect because the mail provider does not transmit it correctly.
If you are sure newsletters are arriving them (not bouncing), then there is no need to “clean” because of open rates. As you already mentioned, it can’t be guaranteed to get a reliable open data because of the mail clients’ nature.
I’m sure the newsletters are arriving, so on that point you’re right — there’s no need to clean it.
However, there are more than 5,000 subscribers on my list, and over 800 (!) have never opened an email. I’m quite certain that’s just due to incorrect data, but otherwise it really ruins my open rate. I run ads from time to time, and the difference between an open rate of 32% and 42% can easily mean more than €1,000. So it’s definitely not something I want to neglect.
Ghost has a notorious issue with fetching data from Mailgun from time to time, as you pointed out as well:
Apart from that, I would try not to rely in open rates for ad reporting in newsletters (independent on whether it’s on Ghost or anywhere else).
Those might be somewhat accurate a couple of years ago, but since many email clients are actively blocking the tracking pixels, these metrics are pretty useless, in my eyes.
I know, easier said than done. But coming from a purely technical side, I’d not trust open rates on either newsletter provider or data sheet.
Edit: some background from people a lot smarter than me in this field
I’ve read this many times before, but these other metrics are much more difficult to extract from Ghost. Sure, I can read click rates, but I’m pretty sure that some of the people who supposedly never opened an email did click on something. Of course, I can’t prove it, but I can’t imagine that 800 people would never click in 6 months. So that would also be a hardly reliable source of data.
In addition, design improvements are often mentioned as a way to increase click-through rates. Well, I would say that Ghost is rather limited in this regard.
I totally understand you, but I can only track a limited number of figures (I can’t even really segment because the (standard) subscribe form doesn’t allow any extra information). I have to translate my metrics into money somehow, and open rates work very well for that. But if nothing is reliable due to all the restrictions, then newsletter marketing is kind of crap.
One more thing: the text you linked to says, “Warm-Up New Email Domains – Sending thousands of emails from a fresh domain can get you flagged. Increase volume gradually.” – Would it be better not to send the next emails with my “new” domain all at once @jannis ?
In Ghost, at least, it is a pretty reliable source, given that click data is “first party data”. Every link you put into a Ghost newsletter has a unique ID that is directly tied to a member. When that links then clicked, Ghost registers the click, updates the database, and then forwards the member to the actual link.
So, from a technical perspective, a lot more reliable than an open event (which is just a pixel that is blocked by most email clients by now).
If you see a correlation here (the people who haven’t opened the newsletter also never clicked on it), I would actually argue that the data is pretty accurate after all?
For context for others: @mynoxin recently changed the custom sending domain on Magic Pages.
With your email volume and given the fact that your domain does have a proven email history, I would not worry about this. I rather see this as an issue with completely new domains, immediately sending emails to tens of thousands of recipients.
I just checked: there are people who haven’t opened the email (i.e., not tracked), but have clicked on it. Thanks for pointing that out. I would never have thought to combine those two (if you click, you have to open it).
Is it possible to select “has clicked at least one email” in the filters? So far, you can only select specific emails.
So: list cleaning is not recommended, but increasing the click rate is.
Adding more links to your post that will encourage people to click will be helpful to get more data about engagement. I think this can be considered as a best practice. Even sometimes calling people directly to read article on browser (by promoting a better reading experience) is a very common practice.
