The publisher report is a day late due to reasons completely out of my control, mainly that the day ended before I sent it out. So here we are at last, me typing words into this announcements section, and you reading them.
I have just published I can't believe I'm about to build another dumb web app to my Tools for Online Speech newsletter. Summary: as I mentioned a few weeks ago, The Sample is a nice side business but has so far not had strong enough business metrics to justify continuing to work on it full-time. I was planning to start freelancing part-time, but... I've now decided to go through with one last pivot. I'm going to build a reading app. If it's successful, I'll integrate it with The Sample. Read the post for more details.
For the next week or two, I'll be automating various processes in The Sample that are currently manual. I've almost finished automating issue approvals: instead of skimming the subject lines every day and blocking the issues that shouldn't be forwarded (like welcome emails, promotions, and the new "X recommended Y" emails from Substack), I've written some code to check for certain conditions and try to filter out emails automatically. In my tests, it works most of the time :).
Then I'll automate newsletter submissions. Currently after someone submits their newsletter at thesample.ai/submit, I get a notification email, and then I fill out a form with their newsletter information, subscribe to their newsletter with a newly-generated @import.thesample.ai, then send them a copy-and-pasted welcome email. I'm going to change this so after you submit your newsletter, it shows you the @import.thesample.ai address right away and tells you to add it to your list yourself.
And finally, one-click subscribes, the bane of my existence. I am glad that I haven't had to do them myself for Substack newsletters since January, although I do wish Substack would release an API for integrations like every other newsletter platform on the planet so that those of you on Substack could stop doing it manually.
Anyway: for everyone else, I'm going to add a few more subscriber-import integrations, including a Zapier integration. Then I'll make it so whenever you get 1-click subscribers, you will get an email that says "hey you got new subscribers, here's the CSV, please import them into your list; and by the way if you don't want to do this manually each time, [click here] to enable one of the integrations to do it automatically."
The one exception will be for those who have enabled paid forwards: I'm happy to keep doing those manually since I don't want to require any workflow changes for people who are already paying us money :).
So, once those things have been automated, The Sample will be a pretty easy service for me to maintain indefinitely. Woo hoo. I'm happy to get out of the "do things that don't scale" phase, and honestly I am going to be a bit more wary about adding manual processes to my applications going forward.
One final slightly disappointing announcement: I'm not sure if I'll be able to keep using smpl.substack.com to do cross-promotion via Substack recommendations. We've received about 100 referrals from it so far, which I am very appreciative of! However, engagement from referrals has been practically nonexistent. Out of all those people, there has been exactly one 1-click-subscribe. (And interestingly, the person who did that sent me an angry email saying in essence "I didn't sign up for this, why are you emailing me.")
Compare that to a group of 56 users who came from a newsletter ad I ran around the same time, which have together resulted in 48 1-click-subscribes. By that metric, subscribers I've received from Substack recommendations are 90 times less engaged than subscribers who sign up via our normal landing page.
So from a practical perspective, it just isn't sustainable to use Substack recommendations for The Sample's cross-promotion program. I'll probably leave it running for one or two more weeks just to see if anything changes, but I might have to shut it down after that. I think recommendations would be better if Substack changed the UI. The current list of checkboxes is too easy to click through without reading (as evidenced by that angry email guy). I think they should change it so each recommended newsletter has its own subscribe button right next to it. If you want to subscribe to the recommended newsletters, you have to click on each one individually.
To be clear, this would cause the number of additional subscriptions from Substack recommendations to plummet. On the bright side, people who do subscribe will actually be engaged, and everyone else won't have their inbox get drowned out by a bunch of newsletters they didn't mean to sign up for. I probably won't mention this to Substack myself since I'm not really a Substack user, but if you are one, feel free to suggest it.
In any case, I'm glad Substack is working on recommendations, even if they possibly have some kinks to work out. In the mean time, if you are using recommendations, I would recommend (ha ha) trying to monitor how engagement from recommendation referrals compares to your other subscribers. This might be difficult since the subscription source just says "Substack Network," instead of giving you something more granular... that would be another feature I'd request from Substack if I was on their platform.