Leon

if you choose the right type of app, there is a super simple/lazy approach to marketing that works I grew one of my apps from 0 to a steady $3k MRR in about 4 months and growing. I am a 10+ year engineer who really sucks at marketing: I chose a traditional industry, and a specific niche within it. It's so niche you rarely ever see posts related to this niche on X. It's basically uncompetitive What didn't work: Cold outreach on linkedIn after I launched a prototype. I have done accelerators before (Antler) and you should obviously just try to talk to people as much as you can, but I found out people in this niche have little clue and seemingly little interest in AI, and are not very open to parachuters Meanwhile what was working: Running targeted google ads. Why? The app is super niche, it's cheap/ uncompetitive bidding, and it's hard to figure who is actually part of the ICP from online profiles, so google ads was actually the perfect fit. Ad spend: I was either slightly above or slightly below breakeven for 3 months of running ads, spending $5k USD/ month. The MRR is on top of that Retention: At first terrible (50%), no surprise to anyone getting their first cohort on a new app. It was so ugly I stopped all ads for a month and did a complete rebuild of the app. This was far harder than getting the thing launched. I’m now < 10% MRR churn for new customers this month and finally getting some real “bravo” kind of emails from customers. An anti pattern: Google ads to get real customers so you can iterate on customer feedback is actually a thing you can really do, even though it goes against all typical accelerator advice. Having 100+ customers coming in from google ads, some demanding demos, some wanting refunds, some excited but wanting help, made it clear what to build and how to present things, very hard otherwise Also: I did free trials on the site for the first month, it didn’t convert well. I put up a paywall, and right away I had new paying customers every day. (also bad about free trials: people forget about the trial, it converts and they never cancel, and this is NOT GOOD to keep collecting that revenue because when they realize they’ve been charged for months for something they haven’t used, you’ll get a vindictive chargeback and this puts you on thin ice with your payment provider. All but one of my chargebacks came from that first month where I offered a free trial.) Tl;dr one winning strategy that's simple: pick a super niche industry, run super targeted google ads. voila

@signulll The door is open for new types of dating app? We need an agentic dating site where it processes every single person on earth to find you a perfect match, and then gives you a link to friend them or something.

it’s remarkable seeing online dating as a product complete its lifecycle. i have rarely ever witnessed products that just mechanically stop working for most of the customers ever before like this, does anyone have a good comparison? these products worked when the userbase was positively selected (early adopters, normal-ish people, 2012–2018 tinder boom) & the algo was quite dumb & they were not monetized to hell. but now selection runs the other way where staying on the apps is itself a negative signal, which is a death spiral no product fix can ever escape. this is pure adverse selection eats the market, aka lemons but for romance. pretty fun to see.

@araseb_ Yep always nervous with new apps to open Stripe or RC, afraid to see cherished mrr churned away

@trenchesbase Leaky bucket syndrom. The worst.

I’m a solo founder, scare me with 1 word.

@ConnorTalksAI @BatsouElef True, it just isn't meant for offering voice cloning as a service, like offering that to my users. Grok doesn't allow you to clone voices via API without an enterprise setup, so really that's the main feature I'm holding out for 🤞

@trenchesbase @BatsouElef There are lots of tools for that like ElevenLabs, hope that helps!

xAI is introducing a Voice Agent Builder: a no-code platform for production voice agents with Grok Voice. Are you going to try it?

@thejustinwelsh Interesting that this is also a luxury you get from having no money.

One of the biggest luxuries money can buy is a week with zero meetings.

@corbin_braun definitely will use it ad hoc. but $10 per 5 minutes means not very often

the endgame model is here. and has been here. the question is can you afford Fable 5?

@levelsio in case people forget, if you hit your Claude code limit Cursor also has Fable. squeeze more while you can

LAST DAY FREE FABLEEEEEEEE

here's a Claude power user move: - get api keys from Replicate and Fal - make a directory call it “everymodel” -->“Claude make a simple app that can arbitrarily generate from Replicate and Fal" - update settings.json or use /add-dir to give Claude "everymodel" for every project every session and project moving forward now has access to every generative AI model from all the providers

@robj3d3 "recreate Adobe Photoshop from scratch"

I have 3 Claude Max 20x plans and 12 hours to burn them all. Give me your best prompts. I'll run them on Fable before it's gone.

@starter_story Time to actually market my ios alarm app 😂

Jake, who built a $50K/month alarm app in 4 months, says the most valuable skill right now isn't building it's... "Focus on top of funnel only. That is the most valuable skill right now is marketing. There's so many people that can build an app and it's so easy with AI, but there's not so many people that know how to market and know how to get views." "If nobody knows about your app, it doesn't exist. You can build as many features as you want." "Building doesn't really matter so much. You got to build a good product, of course. But getting views is more of the unknown."

@athcanft Is he saying he will never be paid for expertise? 🤔

No shade to this person whatsoever, but this will NEVER be me. Fck that. Never.

It proves how deep the proprietary breakthroughs must go. It's like it really happened mainly once: OpenAI & Claude spun out from the same team, China's open source models (maybe, probably ?) distilled from that team's work, and everyone else (Meta/Google/Grok) has to go figure out a crazy amount of difficult things from scratch

Honest question Not a single human have ever spoken to uses Gemini for coding. This seems very strange. Why is Gemini so garbage at coding when Google has scoured the web full of code for years?

@coryalthoff The cost of ads. Not ideal but definitely can work to get your first cohort

early-stage founders: what's the hardest part of landing your first ten customers right now?