From pmf-superpowers
Use when a founder is validating their idea with customers, doing customer development, building a target customer list, running discovery interviews, evaluating warm vs cold outreach, or claiming they have product-market fit. The second Terrifying Question from the Sequoia Arc framework. Resolves only with 50+ cold-outreach conversations on record, <30% warm-intro'd, ≥30% lean-in rate, and a secured design partner with rabid demand.
How this skill is triggered — by the user, by Claude, or both
Slash command
/pmf-superpowers:arc-q2-do-people-careThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
The second Terrifying Question (Sequoia Arc framework, sequoiacap.com/article/pmf-framework-2). Q2 is where most founders rationalize hardest. The pull is to claim resolution from beta-user enthusiasm, warm-intro positivity, or "everyone we talked to said it's interesting." Sequoia's resolution criterion blocks all three.
The second Terrifying Question (Sequoia Arc framework, sequoiacap.com/article/pmf-framework-2). Q2 is where most founders rationalize hardest. The pull is to claim resolution from beta-user enthusiasm, warm-intro positivity, or "everyone we talked to said it's interesting." Sequoia's resolution criterion blocks all three.
Resolution criterion (Sequoia, paraphrased): rabid customer interest, plus secured design partners. (Sequoia article 2: "If you're seeing rabid customer interest and you've secured design partners...")
Where "rabid" is operationalized below.
Requires arc-q1-right-to-exist resolved AND archetype-detection on record. Q2's evidence bar varies by archetype:
Sequoia, verbatim: "50+ conversations in a week" with potential customers.
This is the TDD-for-PMF equivalent. The 50 is calibrated for signal extraction, not market size. Small markets still need signal. Niche verticals still need signal. The rationalization "our market is too small for 50" is entry #26 in founder-rationalization-defense.
Founder-led calls count. PM-delegated calls do not count toward Q2 resolution (delegation is entry #11 + #12 in the rationalization table).
Threshold: 50+ cold-outreach conversations on record before Q2 can be claimed resolved.
Evidence format:
Threshold: <30% warm-intro'd. Above 30%, refuse Q2 resolution.
Why this matters (Sequoia, verbatim): Warm leads are "predisposed to 'be nice.'"
Evidence format: Total cold count / total warm count, with the ratio computed.
Threshold: ≥30% lean-in rate across the 50+ conversations.
Why this matters (Sequoia, verbatim): "If only one in 10 customer conversations lean in, the problem may not matter as much as you thought."
Lean-in is operationalized as: prospect spontaneously volunteers a use-case, asks unprompted about pricing/availability, refers a colleague unprompted, offers to pay or pilot, or commits a calendar follow-up without a chase.
Below 30%: Refuse Q2 resolution. Return to arc-q1-right-to-exist — the wedge or the ICP is wrong.
Threshold: at least one design partner with named rabid behavior.
"Rabid" is operationalized as:
All four. Any one missing → not rabid → Q2 not resolved.
The 50+ conversations only generate signal if they're conducted correctly. The Mom Test rules (paraphrased):
If the conversations don't follow these rules, the 50 don't count. They're 50 polite conversations, not 50 signal-extraction interviews. See founder-rationalization-defense entries #5-9.
For founders in the Future Vision archetype, the signal pattern shifts:
The cohort + the milestone substitutes for the 50-conversation rule.
Q2 Resolution
=============
Archetype: <from archetype-detection record>
Q1 status: <resolved on date, or in-progress>
Conversation log:
- Total cold conversations: <count>
- Total warm conversations: <count>
- Warm ratio: <%>
- Period covered: <date range>
Lean-in metrics:
- Lean-in conversations: <count>
- Lean-in rate: <%>
- Threshold (Sequoia): ≥30%
Mom Test compliance:
- Founder talk-time average: <%>
- Past-spend evidence captured per conversation: <yes/partial/no>
- Compliments-without-specifics rate: <% of conversations>
Design partner(s):
- Name: <named partner>
- Paying/paid-pilot: <amount or LOI>
- Usage frequency: <weekly/daily, named>
- Unprompted referrals: <count>
- Public reference willingness: <yes/no, by date>
Rabid-demand verdict: <pass/fail per partner>
Resolved: <date>
When invoked:
arc-q1-right-to-exist.refuse-below-threshold. Route founder back to the specific deficit.refuse-below-threshold Q3 section (v0.1) or arc-q3-change-behavior (v2+ — currently deferred).DoorDash (Tony Xu, Stanley Tang, Andy Fang, Evan Moore, ~2013). Conducted three or four hundred conversations with small businesses (Sequoia article 2 quotes "three or four hundred") — an order of magnitude beyond the 50 floor. Wrong initial hypothesis: "all small businesses need delivery infrastructure." Discovery moment: restaurants — specifically independent restaurants — had the rabid demand. Refused non-restaurant merchants despite available revenue (entry #17 in founder-rationalization-defense). The Q2 resolution wasn't "all SMBs care" — it was "independent restaurants care rabidly, and we will refuse the rest."
Robinhood (Vlad Tenev, 2022 ICP revisit). Q2 had been claimed resolved earlier (with the original "amateur retail investor" ICP). When the market shifted, Robinhood re-ran Q2 and discovered active traders, not amateur investors, were the cohort with rabid demand. Q2 is not resolved-once-and-done. It can regress when the cohort shifts.
The highest-frequency rationalization zone. Cross-reference founder-rationalization-defense:
arc-q1-right-to-exist (prerequisite).archetype-detection (calibrates which evidence rules apply).refuse-below-threshold (gate enforcer).founder-rationalization-defense (the rationalization table).corpus-retrieval (DoorDash, Robinhood, and more cases of Q2 resolution patterns).arc-q3-change-behavior — v2+ deferred; use refuse-below-threshold Q3 section in v0.1.Generated v0.1.0: hand-authored from canonical sources (Sequoia Arc framework article 2 at sequoiacap.com/article/pmf-framework-2; The Mom Test by Rob Fitzpatrick, 2013, paraphrased methodology; DoorDash and Robinhood cases from Sequoia article 2). v0.2.0 will be regenerated via EnactSkill /generate with full PDF source ingestion.
Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub pauldabrowski85/pmf-superpowers --plugin pmf-superpowers