My article was published on @entrehandbook on @Medium
I've determined three main mistakes that have made an impact on bad output:
1. Wrong tech stack
2. Late fundraising
3. Founders' roles confusion.
Check out the article, and please share your opinion
https://t.co/LYj8oXMIbn
@ChiefSnack Totally agree! I usually also track LTV/CAC for the same cohorts, then you can see not just where retention breaks, but whether the economics still work
I think context is the easy part, anyone can store more docs. The winner is whoever can verify a product decision the way a test suite verifies code, that’s the unsolved part.
If you’re building in this space -- how are you closing the loop, not just the context?
Two weeks ago I asked who’s building “Cursor for PMs”
80+ people replied, 32k+ post reach 🫣
Almost every answer was the same: connect the PRD, sync the tracker, pull the analytics, give the AI memory
A PM’s output is decisions, priorities, and trade-offs. The “success metric” shows up weeks later -- noisy, delayed, multi-causal. So the hard part was never memory, it’s the feedback loop
Not 100% sure, but the feeling is strong
What do you think? Do we actually swing back to brand and offline or is everyone fine with AI UGC and this is here to stay?
#AISlop#BrandStrategy#GTM
I think we're all about to get sick of AI UGC
Everyone ran into it at once because it converted and looked real, built trust
But now the feed is the same faces, the same hooks, the same "authenticity" generated in bulk
And the trust the whole thing ran on is burning out fast
My gut says the pendulum swings back to brand
Not more content, less of it -- but real
The stuff you can't generate -- expensive physical things, offline, a Louis Vuitton trunk in the middle of a city square
When content is infinite and free, the scarce thing becomes valuable
Was just walking around SF and stumbled onto this — a full crew shooting ONE @lyft bike. Softbox, light stands, a truck, the bike propped on a riser on the sidewalk.
@lifeatlyft you're still doing it this way? 😅
Come to us — same on-brand result, AI-generated
@Watchable_ai
I get to be a little bit of a creative director and brand manager again — thinking about the scene, the styling, the talent, the emotional tone, the product moment
I’m honestly very excited about how far generative models have come
Not in the abstract “AI will change everything” way, but in a very practical creative-production way
@0x_nik0@cursor_ai slack channel is basically the workaround everyone ends up with
but it is still messy memory, not structured memory
great for dumping context, bad for retrieval / traceability / why did we decide this 3 weeks ago?
Has anyone already built @cursor_ai for PMs and marketers?
I’m tired of moving project context around via MD files from Claude to OpenAI to Perplexity and back
I want one terminal-style workspace connected to shared project memory, where I can plug in different AI models, keep context persistent, and stop paying for 10 separate subscriptions
Basically: @vscodevz / @cursor_ai, but for strategy, research, growth, product, and marketing