Talk 03 · After the first 10
Scaling Outbound for Founder-Led Sales
past discovery, into demand.
How we went from 10 → 100 customers with cold outbound — the exact PQS + PVP framework, a real example, and the stack you can copy this week.
Audience
B2B founders past their first 10 customers (ACV > $3,000), running founder-led sales, who can already find and reach their customers and are ready to turn the volume knob up without losing reply quality.
About
The closing talk in the series. Same Customer Development discipline as Talks 01 and 02 — but now applied to demand generation. Builds the PQS + PVP framework from Jordan Crawford / Blueprint GTM, walks through Tarka's real PostHog-CDP example end-to-end, and hands you the agent-driven stack to run it yourself.
Key takeaways
- 01You're past 101 — Patrick's Founder-Led Sales Guide is required reading before this. Outbound picks up where it ends, and it's the only channel where you fully control the volume knob on day one.
- 02"One message to 10,000 people" gets ~0.3% replies. The fix isn't a better template — it's a smaller, sharper segment.
- 03Pain-Qualified Segment (PQS): group prospects by a current, data-detectable problem — not industry or headcount. The signal lives in unstructured data and you build it, you don't buy it.
- 04Permissionless Value Prop (PVP): a finished, useful artifact the prospect would pay to receive. Test: could they ship from it today even if they never reply to you?
- 05The Tarka example — startups using PostHog with no custom CDP events. An agent visits each site, inspects network traffic, flags accounts firing only autocapture. The PVP is a custom 5-step instrumentation funnel, delivered as a paste-able skill in a per-prospect Google Doc.
- 06Result: 50% meeting-booking rate across 50 founders from one cold LinkedIn message. Reply rate is the vanity number — meetings is what matters.
- 07Stack: evergreen lists (Product Hunt / Crunchbase / sites with incomplete tooling), Claude Code does the qualifying + enrichment, HeyReach delivers connection-request + Doc-link on LinkedIn, allman.sh handles replies once you have enough past conversations to train on.
The rest of the series