@willccbb@franciscojarceo Post-training a model is cheap. What’s costly is maintaining the model. Do you think these companies are ready to make this long-term bets?
"Never scale your team until you have true product-market fit."
Honglei Liu (@ppuliu) - Co-Founder and CTO of @Tofu_HQ & former AI researcher at Meta—explains the lessons of the collapse of Fast (raised $100M+ before shutting down) and the playbook of finding PMF.
Here are 5 brutal lessons for early-stage builders: 👇
• 𝗣𝗿𝗲𝗺𝗮𝘁𝘂𝗿𝗲 𝗦𝗰𝗮𝗹𝗶𝗻𝗴: Hiring before PMF kills focus and slows pivots.
• 𝗙𝗼𝘂𝗻𝗱𝗲𝗿 𝗧𝗿𝘂𝘀𝘁: 1.5 years of working together—even at the same startup—to validate alignment.
• 𝗖𝘂𝘀𝘁𝗼𝗺𝗲𝗿 𝗗𝗲𝘀𝗶𝗴𝗻: 50+ CMO interviews; ignored polite feedback, chased “I need this now.”
• 𝗣𝗶𝘃𝗼𝘁: SMB churn → enterprise demand for integrated AI marketing.
• 𝗠𝗶𝗻𝗱𝘀𝗲𝘁: Big pay cuts to build a lasting B2B system, not a point tool.
@JayHoovy Harder said than getting done. I asked -30 founders in the past week and most of said distribution is at the top of their mind. But hard to get it right.
@daxeshpatel the 47% gap is real but not because of tooling. it's because demand gen still optimizes for mql velocity and abm optimizes for deal probability. you can't fuse two things that have different success metrics without someone losing their bonus.
Asked ~30 startup founders their biggest challenge over the past two days.
One answer stood out: go to market.
Some want to accelerate the growth they’re seeing.
Some are still hunting for their ICP and how to reach them.
Some have adoption in their base and want a new segment.
Some want to turn founder-led growth into a repeatable playbook and a GTM team.
Feels whoever cracks repeatable GTM for startups will have outsized impact.
(full disclaimer: I actually think startups are the worst customers. Eager to try anything, always pivoting, never sure what they actually need. 😂)