loading…
RoastMyPitch itself: B2C consumer product where founders sub
v149d5.4GO · REFRAME
5.4/100Medium-High confGO · REFRAME
“Self-aware meta-pitch with real engineering, zero commercial proof.”
Marla Chen5.8“Clever self-roast, but ops math is still hand-wavy.”
Devon Ruiz4.2“TikTok CAC target with zero proof is a vibe, not a motion.”
Rita Patel6.1“Clever frame, but the moat lives in the panel characters.”
Jax Okafor5.8“Meta pitch, real self-awareness, shallow monetization moat.”
Top 3 action items
- P0Get five non-founder strangers to pay for Pitch Coins
- P0Build the unit economics on one roast
- P1Reframe the product around the panel as recurring IP
Marla ChenPanelClever self-roast, but ops math is still hand-wavy. One engineer plus AI pair is a real answer to 'who's writing the code' — I'll take it. The prompt-stability risk is named and mitigated with pinned versioning and golden tests; that's exactly the kind of assumption-labeling I want to see. What's missing: the unit economics on Pitch Coins are gestured at ('~150 bundles') but there's no stated price point, COGS per review (Claude API cost per roast run), or gross margin. Sub-$10 CAC via TikTok/X is a hypothesis, not a model — what's the assumed conversion rate and LTV that makes that number work? With 1 user and no revenue comps, the premium tier with ex-YC consultations is a revenue line on faith. Show me the ops math on API cost per roast, then we can talk.
Devon RuizPanelTikTok CAC target with zero proof is a vibe, not a motion. ICP is founders — okay, that's a named segment, I'll give you that. But 'TikTok/X with sub-$10 CAC' is a hope, not a channel plan. Where's the content angle, the hook format, the posting cadence, the test budget? You have 1 user — yourself — which means zero signal on whether any acquisition channel works at all. The Pitch Coins revenue model is plausible but secondary; I need to see five non-founder strangers pay before I believe the $10 CAC is real, not a back-of-napkin number.
Rita PatelPanelClever frame, but the moat lives in the panel characters. Strip the AI personas and this is a feedback marketplace — which already exists (MicroAcquire advisors, YC office hours, Reddit). The frame that survives reframing is 'the panel IS the product': Marla, Devon, Rita, Jax, Judge as recurring characters founders want to impress. That's a genuine insight. But the pitch doesn't lean into it — it leads with 'AI roast' which collapses to 'ChatGPT with a mean prompt.' The Pitch Coins model and ex-YC consultation tier suggest the founder suspects the real value is human signal, not AI signal — that tension is unresolved. Pinned panel_v1 + golden tests shows the founder has already thought about prompt drift, which is exactly the kind of rejected-weaker-framing work I reward.
Jax OkaforPanelMeta pitch, real self-awareness, shallow monetization moat. The founder is pitching the tool they built to roast pitches — that's either clever or recursive cope, and I'm genuinely unsure which. I've seen 'AI feedback marketplace' before: the graveyard includes pitch-deck scanners, AI co-founder tools, and YC rejection analyzers, all of which died because founders want validation more than truth, and truth doesn't retain. The Pitch Coins model is a real attempt at friction-as-filter, which is smarter than freemium, but the CAC math on 'sub-$10 via TikTok/X' for a product with 1 user is a hope, not a thesis. What saves this from a lower score: naming the technical risk explicitly (prompt stability, pinned panel_v1, golden tests) is exactly the kind of graveyard-awareness I reward — most founders in this category skip straight to 'AI will figure it out.'
No human replies yet.