| Hey, How can you identify which readers are truly disengaged, and what can you do to remove them safely? These are the questions we will answer here today. In case you missed it, we've been talking about list optimization over the last couple of weeks; you can catch up here: It's Hard to Let Go. And for everyone emailing me about Substack – which is quite a lot of you, given the sticker price – we're discussing that over on my Facebook Page this week, where you can read what I wrote about it a few years ago, as well as more up to date experiences from a range of writers. Please share your experience too! the numbers Before I show you how to optimize your list, let me share data from two recent list optimizations I did – as they will illustrate the benefits better than any argument. The first is a list I optimized quite recently, approximately 18,000 subscribers. - 60.39%
- 58.66%
- 56.78%
- 56.68%*
- 63.2%
- 65.77%
You can see the clear increase after the list optimization ran (marked *), improving open rates by almost 10% to what is an extremely good open rate for a list this size. Here are some recent open rates for an even bigger list I manage - approx. 25,000 subscribers - which I optimized a few months ago so I've even more improvement to display: - 54.28%
- 57.19%
- 55.18%**
- 71.94%
- 72.07%
- 69.03%
- 74.29%
- 76.31%
In this case the open rate increase after the list optimization (**) was even more significant, improving by over 20% to what is a really amazing open rate, especially for a list this large. The first list is a weekly non-fiction newsletter BTW and runs optimizations maybe once a year. The second is a monthly fiction newsletter which does this process twice a year on average. Both cadences are fine, to me, for whatever you happen to write. In the first scenarios the only subscribers flagged for potential removal were those that hadn't opened or clicked anything in over three months, and in the second, six months. So you can adapt as you wish to your circumstances. It's pretty light touch list optimization either way, as far as I'm concerned, and all that most authors will need to engage in unless they have serious problems with their newsletter, such as high spam reports or plunging open rates that they just can't turn around. how to optimize your list A lot of the arguments around list optimization are silly because the opposing side seems to assume that we just go in and delete people. A good process is not so reckless. As I said last week, we need to: - Determine which readers are truly disengaged.
- Construct a sequence which will attempt to re-engagement.
- Word the email carefully (and leave the door open for return).
- Tag people so you can keep track of everything.
Open rates are a little squishy these days for a whole range of boring technical reasons so it's best to use them in combination with click data (much more solid) to determine who actually is disengaged. One of my preferred providers – Kit – has a really nice way of visualizing engagement across your list, and most providers do some variation of this, although not always with the nice visuals. Here's what this list looks like right now: This is pretty much the ideal – which is what I would hope to see directly after list optimization. 83.34% of my list are deemed 5-star subscribers by the system, meaning that they have opened or clicked something in the last 30 days. That's an excellent level of engagement and about as good as it gets on a decent sized list – in my experience, at least. A further 10.8% are considered four-star subscribers – i.e. they have opened or clicked something in the last 30-to-90 days. That makes a total of *math noises* 94.14% of my list who have opened or clicked something in the last three months. I'm really happy with that; it's a very engaged list. The numbers were moderately less impressive before I removed those who were likely disengaged, but how did I determine that? In MailerLite – my other recommended provider (see last week's email for a comparison) – calls them Inactive subscribers, whereas Kit dubs them Cold subscribers. I prefer the former term as the latter is often used to distinguish between organic and inorganic subscribers, so I'll go with MailerLite on this one. Whichever provider you use, one of these two or another, there should be either some automatically determined segment which are candidates for removal, or some way of creating your own cohort. Every provider is different, but most have some way of slicing and dicing your database so that you can come up with your own list – and perhaps I prefer doing that than relying on what a provider thinks is a disengaged reader. You can decide how aggressive or conservative you wish to be here. Some people go super aggressive (YMMV) and boot people who don't make it through onboarding without clicking something, for example. Or toss those who go sleepy for a month. And sometimes have automations constantly running to manage all this. I'm much more conservative and hands-on here. I only tend to optimize my list once or twice a year and prefer to do it manually. I usually make a group (or tag or segment, depending on your provider) of subscribers who haven't opened or clicked anything in 6 months. Anyone in this group, who is not tagged with NEVERCULL – more on this in a moment – will move forward to the next step. stay/go Opponents of list optimization also don't seem to realize that we don't simply delete readers who have gone dark – we attempt to bring them back into the light first. No matter how careful you are, or how solid your data is, some people will slip through the net; here's where we catch them. Before removing anyone from your list, at minimum, you should send them a Stay/Go email. Which is exactly what it sounds like – an email asking them if they wish to remain on your list – but execution is everything here. Two aspects particularly demand your attention: (a) The email should be as pared-back as possible to maximize deliverability. You are emailing a tranche of your subscribers which is not opening your emails so it's safe to assume that a large portion of those people simply aren't seeing them, because they are dropping into Spam or Promotions, or perhaps not being delivered at all. In practice, this means we should act like the recipient's spam filters are incredibly sensitive and absolutely minimize the biggest triggers – things like images, links, and spammy words. I typically put no images in these Stay/Go emails, no spam words in the subject line, and only one link (the "Keep Me" link), as well as the obligatory unsubscribe link in the footer, of course. (b) The email should be worded very delicately. Some readers can take this email quite personally. Perhaps they haven't been getting your emails for whatever reason and this message is a bolt from the blue which might feel rude. Or maybe they are going through something which has caused them to let their inbox slide. Whatever the reason, I've found that some readers can be quite sensitive to a Stay/Go email unless it is worded carefully. I like to use lots of passive voice and blame the system, man, the aim being not to inadvertently blame them for not opening my stupid emails. It's also important to keep the door open: readers can – and do – return over time. In perhaps surprising numbers. I always ensure people know they are welcome back any time. I usually set this up as a mini automation which will do the following: - Send a Stay/Go to anyone with the list optimization tag.
- Remove anyone from the process who clicks the KEEP ME link (usually a big red button).
- Tag anyone who clicks with NEVERCULL so they don't get bothered by this ever again.
- Unsubscribe everyone who doesn't click KEEP ME after 7 days.
It can also be a good idea to tag those who you do remove, so you can track who has returned to your list later on down the road. (I also like to manually add these to NEVERCULL so I don't remove them or annoy them a second time.) The above is a fairly basic re-engagement/removal process. You can attempt a lot more re-engagement before removal – and many authors do. For example, this can be a great time to engage in a little old-fashioned bribery. You can dangle your magnet again, or some other exclusive short story… if they decided to stick around. Or you can simply tease upcoming content, or hype an impending release. It's really up to you and authors handle it in all sorts of different ways. Something else you might consider: someone might still want to follow you but not receive your emails, so it can be a good idea to suggest they follow you on Facebook or your network of choice instead. Of course, there's a trade-off here – you are putting in an extra link meaning you might not reach quite as many of those more sensitive recipients we spoke about earlier. But this is where monitoring your data can be helpful. If you see no one is really clicking those links, then you know there's no real upside to including them. to automate or not? While there are many aspects to consider here, the rewards are considerable – as you can see from my numbers up top. Yes, it's a bit of a faff setting all this up, but because I have it set up as a mini-automation, each subsequent time I want to run it, I only need to switch it on or off, and do a five-minute check on my segment to make sure everything is as it should be. And then I can optimize my list whenever I like, with minimal work. You can go further too and have it running constantly in the background without any further intervention from you, automatically identifying when someone goes dark for six months, or your interval of choice, and running them through your preferred re-engagement process – but, as I said, I prefer keeping my hand on that particular tiller. Run your ship as you wish, cap'n. Dave P.S. Writing music could only be Sly & The Family Stone with If You Want Me To Stay. |
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.