Cross-platform recommendations
I've finished working on Yakread's discovery features. (Or at least, the first draft of those features has been completed.) I've also updated the landing page accordingly:
In other words: the default, path-of-least-resistance flow is that you simply sign up to Yakread and then we start recommending stuff. You get 5 links, you pick one to read, you get 5 more links, etc. No effort beyond that is necessary to get value out of Yakread. Articles in Yakread even have a subscribe button, if we were able to find an RSS feed for the article:
So this path-of-least-resistance flow even has a way for you to build up your list of subscriptions.
And then totally optionally, you can add your bookmarks/subscriptions/ebooks from other places.
As such, my gut feeling is that Yakread is in a pretty good place now. I might start booking a few ads for it at least. Maybe I'll start doing "marketing Mondays": have at least one day per week dedicated to growth, and then spend the rest of the week working on feature requests and such.
Speaking of feature requests, a couple that I'm planning to do this week include:
- Add an Instapaper integration for importing bookmarks.
- Let users upload PDFs. (The formatting won't be anything fancy; I'll either embed a PDF reader or I'll just have an "open PDF" button that loads it in a new tab/app. But at least they'll get mixed into your Yakread recommendations.)
I also would like to add highlights, notes, and probably a Readwise integration. BUT before I get to that, I've had what is known in technical terms as an epiphany.
You know how Substack has its fancy recommendation thing, where Substack writers can recommend each other, and this is driving a bunch of subs (of unknown quality perhaps, but still), and it's making people switch from Ghost?
I figured out a practical way to replicate the recommendations feature across different different newsletter platforms! I wrote up some rough thoughts about it, but the takeaway is that this week I'm going to [hopefully] start working on some sharing features for Yakread. You'll be able to paste in links to whatever newsletter/blog you want to recommend—whether it's on Substack, Ghost, Beehiiv, or wherever—and Yakread will give you a link to a subscribe page you can share. If you want, you'll even be able to embed that subscribe page onto your own website and use it as your main subscribe form. It would give you the same flow as a Substack signup form: first people sign up for your newsletter, then they see a list of other newsletters you recommend that they can sign up for.
There'll also be statistics and stuff so you can see how many subs you've referred to other newsletters and how many have been referred to you. A big reason why Substack's recommendations have been helpful is because they've been able to help set norms/culture/encourage people to recommend each other. I'd like to build a similar culture into Yakread, where people share the things they read.
So these sharing/recommendation/subscribe links will be a big part of that. After that's set up, I'd also like to make it easy to take articles you've thumbs-upped on Yakread and generate a newsletter post template from them. For the past month, I have been using my personal newsletter to share articles I've read on Yakread. At a minimum, I think it'd be swell if Yakread had a page where it would take all the articles you've liked in the past week, generate a post similar to the ones I just linked, and provide a "copy to clipboard" button. Then you can paste it into your newsletter platform of choice and make further edits from there. (This'll be extra nice when Yakread supports highlights and notes—we can stick those in the generated post too, in which case you might not need to do much editing at all!)
One other fun thing I've been trying out: I've started using the TFOS forum as a place to record public notes about ideas I have. I already linked to one thread about these cross-platform recommendations, and just now I typed up another one about outsourcing Yakread's ranking algorithm.
Published 10 Oct 2022