I'm thinking about the following strategy: 1. stop doing paid acquisition for Yakread, for now. The current product perhaps just requires too much investment from new users for paid ads to work out. 2. instead, focus on organic growth. continue writing the weekly newsletter, and when relevant post them to HN. Focus on HN as the main growth channel for now. I don't want to sweat it too much--as discussed, psychologically I'm enjoying getting out of the product growth mindset for a bit. But I'll at least keep HN in the back of my head as I write the newsletter. 3. while doing that, continue to refine the product. Honestly if it's geared towards HN, I may want to rejigger the whole thing to focus on RSS. update the landing page, and have the onboarding flow be "upload your opml file." 4. When that whole thing is going well and I want to expand beyond the HN crowd, put more work into yakread's discovery system. do the "tiktok/youtube for reading" thing (https://tfos.co/p/tiktok-for-reading/). i.e. the landing page mostly describes the product as "sign up for this thing and we'll show you interesting articles to read." you can still subscribe to specific newsletters and feeds and things, but it's not emphasized on the landing page or onboarding flow--it's more focused at people who have already become active yakread users and now want to customize their experience some more. 5. At that point I think paid ads would work well.

Some interesting stuff going on here https://hq.getmatter.com/your-new-inbox

As I've been pondering #1095430452145299517 more deeply, I'm wondering: instead of just tweaking a few parts of yakread to make it amenable to hacker news users, what would it look like if I reimagined yakread from the ground up for that use case? Reiterating the background: this is interesting to me because the marketing channel I was banking on--paid newsletter ads--I think won't actually work for me, at least not yet. Hacker News as a primary marketing channel I think could work though (especially since unlike e.g. twitter, I'm actually pretty good at writing content for hacker news). I'm thinking about this also because the latest article I posted to hacker news actually got a handful of upvotes and resulted in 5-10 signups for yakread (https://tfos.co/p/rss-is-great-for-recommendations/, not exactly a high-effort post either). So HN's just on my mind. So I'm imagining a funnel that goes something like this (off the cuff, I'm coming up with the details as I write): 1. I write posts like the one linked above: catchy title (basically just make sure title includes "RSS"), topic is some interesting TFOS-related thing that's relevant to yakread, and there's a link to yakread after a few paragraphs. 2. people on HN read the article and click through to Yakread's landing page. 3. immediately on hitting the landing page, you get a feed of articles from HN. no need to sign in. 4. some features are gated behind account creation. 5. after user creates account, then they get the daily digest email, and maybe I focus more on paid subs and try to upsell people to that (sort of have a feeling that might be more effective with the HN crowd than ads, unless writers are interested in paying for RSS subscribers... I would wouldn't be, personally.) If this worked, I'd run with it for a while and eventually expand it to once again be marketed as more of a standalone thing rather than just a client for HN.

The conclusion of my stream of consciousness in #1096292816969281556 : I'm probably going to rework yakread to give you a default algorithmic/discovery feed right away instead of trying to push people to add their own subscriptions immediately. That will vastly compress the onboarding funnel and will hopefully make marketing/growing yakread much more feasible. biggest changes I have in mind: 1. I'll probably get rid of the "curated"/"recent"/"discover" tabs--I think better just have the one algorithmic feed for simplicity. The daily digest will still have the "recent" section. That matches how I use yakread anyway. probably will keep it how it is for current users, at least for a while. that's actually how the feed was before. 2. I'll make the app usable without logging in. instead of seeing a landing page, you'll immediately get a feed of articles. If you try to subscribe to something then you'll get prompted to create an account. and on the backend, a bunch of algorithm work (especially on discover recommendations). Other than that stuff, I think the app will stay pretty much the same as it is now. Subscribing to newsletters / feeds won't be pushed during onboarding like it is now, but the subscriptions page will still be there and stuff. I may or may not actually focus on HN. probably will try out paid newsletter ads some more and see if they look feasible this time around. if not then I'll focus on HN and make the discover recommendations pull from HN or something. monetization can probably stay the same--leave it as pay-per-click ads.

Some interesting points/theory about the substack vs twitter thing: https://www.thediff.co/archive/dueling-over-platforms/

It's too early to say for sure, but I think I've figured out a way to organize my todo lists that actually fits how I work and doesn't require too much effort to maintain and stuff. Instead of organizing the top level by project (e.g. a "yakread todo" note, a "biff todo", a "writing todo" note etc), I have two fixed categories: main and alt. Main contains stuff that feels "high leverage", regardless of what project it falls under. Alt contains stuff that probably won't have a big impact in the short-term but that I still want to make some progress on. So far (about 2 days) this has made it feel easier to focus on the essential stuff and not have my main todo list get crowded out with too many items. I try to keep each list to 7-ish items, and each item can branch out to a separate note with 7-ish items etc. So for any given note I can prioritize the tasks without too much effort. I have found that using google keep for the notes is essential so that I can organize them on my phone.

I'm rewriting the entire UI from scratch. super secret work-in-progress here: https://yakread.com/home (notably, you don't need to be signed in to go there)

Ad business models from a macro economics perspective https://www.thediff.co/archive/why-the-answer-is-so-often-ads/

Mostly just a hypothetical thought experiment https://tfos.co/p/what-if-i-started-a-nonprofit/

Any idea when the sign-in page will be fixed? I’d like to update my profile and access the table to connect with other writers.

My last post (https://tfos.co/p/what-if-i-started-a-nonprofit/) sparked a couple conversations and has me thinking about what kind of career paths and supporting institutions I wish existed. I think I can describe my main thought in one sentence: those who want to should be able to spend up to 1/3 of their working time on exploratory work. "exploratory work" could be lots of things--entrepreneurship, research, open-source, invention, learning--the defining characteristic is that it's self-directed and not beholden to any external measures of success like profitability etc. I think this should be a thing throughout all of life, starting in grade school, continuing in college, and throughout your career. (In fact for some people it could even be a replacement for college.) As your exploratory work matures, you may be able to transition to working on it full-time. Perhaps it's turned into a profitable business, or you have grants/sponsorship, or a company hires you to keep doing what you're doing. But no matter how early-stage your work is, you can always spend at least 1/3 of your time on it, even without an arrangement like this. It would be helpful to have institutions that (1) help people get a sufficient income with no more than 2/3 of their working time, (2) provide/help people find a supportive community for their exploratory work.

since I've gotten into a groove of writing off-the-cuff-ish thoughts here in discord, I thought maybe worth experimenting with mastodon again and substack notes. I'll just be copy and pasting stuff from here to there: https://substack.com/profile/113327141-jacob-obryant https://social.linux.pizza/@obryant666 While I'm not concerned about trying to get lots of subscribers, I do still want to meet new people and develop relationships and stuff. So I'm thinking about if/how I want to use social media to that end. I think I may need to dedicate more time to looking for people I'm interested in following, and then engaging with them wherever they happen to be writing.

related to #1100875505701486612 . My brain has a thinking side and a building side, and the building side has been screaming to actually get some work done these past few days. But, one last thought... in trying to envision the ideal state of social media etc, I've often wondered if it's necessary to preserve the model of twitter et. al. where you have a bunch of short-form posts. It simplifies things a lot if you work in larger chunks, like a weekly newsletter. That's been an attractive part of the unbundled model where you have newsletters/blogs for broad dissemination of ideas, and short-form posts are kept within small communities like discord servers, slack groups etc. And yet--I have a nagging, half-formed feeling that the simplicity of that model just won't cut it. I think I may need to figure out what the best way is to accommodate short-form posts. I did grapple with that idea to an extent in https://tfos.co/p/rebuild-social-media/. I don't think the ideas there are sufficient though. Aggregating users' posts from various discussion platforms has significant barriers to mass adoption because (1) plenty of platforms don't even have APIs/RSS support, e.g. substack notes, (2) even if they did, the majority of users simply aren't going to connect any external services. In the long-run, it might be worthwhile to add posting capabilities to Yakread, while preserving the ability to use other publishing/discussion platforms for those who choose to. So I'm starting to mull over what that might look like.

As I work on #1099439275348602990, I think I'm closing in on the "right" structure for Yakread to take as a reading app. I plan to limit the scope to reading for a while as I focus on growing the app. But in the end game, IMO it does make sense for the core social media functions--reading, publishing, discussion--to be bundled into a single app. The key is that those functions should still be interchangeable. e.g. you can read blog posts in the same app that you use to publish your own blog posts, but (1) you can still read blog posts that people publish with different apps, and (2) other people don't have to use the same reading app as you in order to read your blog posts. This way you get the fully integrated social media experience by default, which is important for mass adoption, but you aren't stuck there. Power users with different needs can switch out parts of the app for 3rd-party implementations, and the ecosystem as a whole can be more gradually adaptive instead of lurching from one VC-funded behemoth to the next. (Aside: this does have some slippery slope issues, because even if your all-in-one social media app does work with external content, as your userbase grows, the temptation to favor in-network content--e.g. by boosting it in the recommendation algorithm--will always be there. That'd be another benefit of #1100165507509911705. It would be great if there was at least one neutral, open-source social media app of this kind.)

The weekly newsletter: https://tfos.co/p/how-to-deal-with-short-form-posts/ A continuation/summary/refinement of the ideas in #1100875505701486612 and #1101230902513766452

An interesting article about the "unambiguous failure" of social media https://coryd.dev/posts/2023/we-lost-the-thread

It happened again--I wrote another post. https://tfos.co/p/yakread-rss-recommendation/

There's still a bunch more stuff I need to do before switching the new UI on for everyone, and but today I finished adding all the main functionality, so feel free to try out "Yakread 2.0": https://yakread.com/home

tada: https://yakread.com still have some things on the todo list, but I finished enough stuff to make it ready for release. let me know if any thing's broken.

now that yakread is mostly where I want it in terms of features, I've been thinking about what marketing channel I should focus on. I want to take a break from paid ads for a bit and probably try them again after yakread has some organic growth. I don't think seo is a good fit. hacker news comes to mind since I've had a number of articles hit the front page there in the past. though I've also thought a little about substack notes/substack in general. it's appealing since it's a network focused on people who read/write blog posts which is exactly who I want to target. I've just not been sure how I'd go about it. as an overly technical founder, I'm fairly marketing-impaired, and I need a strategy that I'll actually be able to execute on. about 10 minutes ago I had a flash of inspiration: I can funnel my irritation for substack into marketing for Yakread. make a branded substack account for Yakread and then write blog posts + notes that critique them etc. it'd be in good faith--I want all my critiques to be valid/fair--but I will "dish it out" as much as I can. with humor. and it's perfect because I can be completely blunt about how the whole point of the publication is to promote yakread (which competes with substack), and thus I can include plenty of signup-for-yakread CTAs and have it be funny also the content would be relevant to anyone on substack, and anyone else who has beefs with substack would likely be down to share it and this is a topic which I can probably write plenty of stuff about lol. every time substack publishes another PR fluff piece it'll give me more stuff to write about 😈

I just renamed the "Share" button to "Comments." In addition to the sharing links, it now searches Reddit and Hacker News for any discussions of the current article. It also checks the original article for any text of the form

[number] comments

, which works for comments on Substack posts at least. Kind of a small update, but an important conceptual piece nonetheless, as I mentioned in last week's post (https://tfos.co/p/yakread-pre-release/)

Just read https://www.cybercultural.com/p/readwriteweb-20 written by a well-known tech blogger guy of whose existence I only recently became aware. Has an interesting tidbit about the blogosphere vs. newsletters today: > 90% of the battle with email newsletters is getting readers. Large networks matter in newsletters; moreso than in blogging. When I was a young blogger, I made a name for myself by catching the attention of more experienced bloggers — either by linking to them (which they would see via the trackback system) or by commenting on their blogs. By comparison, it’s difficult to gain the attention of other newsletter writers; not least because linking has gone out of fashion. It's triggered me again to think about how to help my writing reach people who would want to read it. #1111072618708213830 is one approach -- try to get good at the social media game (substack notes in particular), use that to drive yakread signups, and then let yakread help distribute my writing at TFOS (I can give myself free advertising credits there after all, and will probably give myself a boost in the recommendation algorithm too). plus then anyone who comes across TFOS will already have some context on my work. And getting users for yakread is the more important business goal anyway; if some of them happen to get into TFOS also without me having to anything extra, then great. A downside of that approach is that I'm really not a social media person. My irritation with Substack might provide enough fuel for me to be successful with yakstack anyway, but either way I'm still interested in other non-yakstack ways to network.

I made a subreddit to experiment with as a possible replacement for this discord server: https://www.reddit.com/r/tfos/ See this post: https://www.reddit.com/r/tfos/comments/13snpvd/tools_for_online_speech/

The weekly newsletter from you-know-who https://tfos.co/p/exploring-reddit-essays/ Also on the new subreddit: https://www.reddit.com/r/tfos/comments/13w3m8a/exploring_reddit_essays/

So I discussed in the newsletter couple weeks ago ( #1113226876735987806 ) that I was thinking of replacing #1048824705131499560 with the new tfos subreddit. However current events have me questioning that decision. I don't have strong opinions about whether or not Reddit is acting badly in wanting to put more restrictions on its API--it's a bummer obviously, but not necessarily a bigger bummer than Reddit being unprofitable--but from what I've heard it sounds like they're handling the situation terribly. I'm concerned about the stability of Reddit going forward. On top of that, even though Reddit does technically have a real-time chat feature (the "lounge"), no one's ever gonna use it. It's just not a common behavior; Reddit is for threaded discussions. That was leading me to think "maybe I should stick with #958102014733189191 here on discord for unstructured chat, and then use reddit just for async discussion." But that really seems like a much less good situation than simply sticking with #1048824705131499560 here on discord, so forum and chat are in the same place. So, my current thinking is that at some point I really need to just polish up my discord forum publisher thing (https://discord.tfos.co/) so that I have publish links I can share in the newsletter. Then problem solved I guess 🤷‍♂️

Just sent this out: https://tfos.co/p/reneging-on-reddit/