Marcus

if you test it, i'd love one honest note on what feels off that’s more useful than a like

if you test it, i'd love one honest note on what feels off that’s more useful than a like

we hit $1M MRR in 30 days from building in public! that would’ve been a great tweet ... but also complete nonsense what actually happened is way less cinematic me and a friend have been shipping a lot lately we have plenty to share but still kept ending the day without doing

would genuinely appreciate any/all feedback:

most builders i know are so focused on shipping they completely forget about content but content is the highest leverage distribution channel available right now if you're building in public and not posting about what you ship, you're leaving a massive amount of visibility on

you pushed to github last night and went to bed without telling anyone... again. SmashLanding reads your commits and writes the post for you, in your voice.

you need to try this

if you want to see what i built:

i wasted SO MUCH time building “cool” AI tools nobody needed. this time i did the opposite: > mined real pain points from operator threads > picked one recurring bottleneck > built only that workflow the bottleneck was local seo agencies drowning at month end reporting. not

here’s what i built:

i stopped building random AI tools and did this instead: picked one painful niche read real operator complaints for days built only around repeated pain shipped that’s how i built GeoAuditKit (local SEO + AI discoverability reporting) what actually worked: 1. pain first

here’s the product:

i shipped a SaaS for local SEO & AI Discoverability agencies because i kept seeing the same thing: great fulfillment weak reporting client churn anyway agencies aren’t short on data. they’re short on clear “what to do next” reporting, especially now that clients ask about ai

a 23 year old with 6 months of experience made $8,700 last month... not from content or selling prompts. he built AI agents for dentists and plumbers. they were already paying $1,500/month for the services he replaced. he just automated it.

cursor has been ignoring your .cursorrules file. you just didn't know. .cursorrules is deprecated. the new system is .cursor/rules/ one .mdc file per rule, not everything in one place. but even after switching, rules get buried around message 40-50. context has a token

the most useful dev content gets the least reward someone spends 4 hours breaking down a solid build decision …gets 12 likes someone posts “10 AI tools you need” …gets 40k views that mismatch is exactly why bootstrappr is interesting it’s built for builders who build and

rip designers hello nano banana 2

I can now leave Codex running, go make dinner, and come back to a working app. That sentence sounds fake until you see what changed. I didn’t find a magic prompt. I changed the way I prepare. Before: • random idea • open editor • prompt while building • patch, reprompt,

wasted 4 hours on google auth today. building with clerk + supabase. what took time: • oauth credentials in google cloud • redirect uri mismatch fixes • identity mapping so db permissions actually work if you’re shipping soon, do this before building features: 1. define

a startup in india just hit $100M ARR. in 8 months. no VC. no senior engineers. just: describe what you want → AI builds it → ship. lovable did the same. $200M ARR in 12 months. together that's $300M in annual revenue built by people who couldn't have written the code