I asked 10 of YC’s fastest-growing founders how they got customers (everything I learned)

Executive takeaway

The video claims YC’s fastest-growing companies follow a nearly identical, “scientific” 4-step growth playbook for 2026—mostly by:

• Choosing easier markets
• Getting extreme customer insight
• Using highly targeted content as algorithmic distribution
• Converting “lurkers” via hyper-targeted warm outbound

---

The 4-step growth playbook (for 2026)

Step 1: Pick a “less competitive” fight (easy market/category)

Fast growth is framed as coming from targeting large, old, hated, and under-digitized industries and rebuilding them with AI—rather than competing head-on with software/SaaS companies selling to software builders.

Framework / selection principle

• “Don’t compete with the best” (avoid the most saturated “software-for-software” categories)
• Target industries with major scale and legacy processes (e.g., fax-era workflows)

Concrete examples

• Corgi (YC): Insurance (described as “boring/hated”)
• Lagora (YC): Legal-related category (described as having little competition)
• Candid Health (YC): Medical billing / denied claims / insurance
• Areni (YC): A narrow wedge—reception desk for dentists’ offices; ~200% growth this year

The video’s argument: these companies often have bigger reachable markets than “familiar” software-for-software YC brands.

Notable “scale” claim / benchmark

• Corgi: 0 → $40M/year in ~2 years, projecting a 10x increase from that point.

Market sizing factoid (used rhetorically)

• Insurance is said to be ~12% of GDP, framed as 2x software’s size.

---

Step 2: Get pathologically close to customers until “fog clears”

Customer research is framed like a “Battleship” fog-of-war problem: until you talk to many customers, you can’t actually see where value creation sits.

Process / playbook

• Talk to hundreds of customers; accept many “misses”
• Build intuition for:
  • the customer profile
  • where problems are
  • what’s valuable vs. not
• Aim to become expert in the problem space, not just your product

Tactical examples

• Sim (Emir):
  • Used a landing page + early product, with a booking link
  • Spent months, ~10 calls/day
  • No “qualification” on calls—just talking with curiosity about problems
  • Outcome claimed: 20,000+ GitHub stars and “viral” growth driven by a unique solution discovery loop
• Origami (speaker’s company):
  • “First 3–4 months”: calls all day every day
• When direct calls are hard:
  • Physically go sit in with customers (watch work, talk on-site)
  • PlanGrid (Tracy, YC): showed up at construction sites with donuts to sell to construction workers—food as an “entry ticket” to attention

---

Step 3: Win with content your customer profile can’t scroll past (algorithmic targeting)

For B2B growth in 2026, the video argues:

• “Going viral doesn’t matter.”
• What matters is conversion quality: content reaching the right buyer segment.

Core claims

• Platforms increasingly recommend based on a 4U-style interest graph (not follower count)
• Followers don’t matter; the algorithm serves posts to the audience matching the content/theme
• Content must stay hyper-specific; off-topic posting “throttles” reach

KPIs / conversion math mentioned (speaker examples)

• Speaker’s own content outcomes:
  • Post with ~2M impressions → 0 revenue
  • More targeted post: ~100k impressions → ~$6,000 MRR (adds in a day)
• Rule-of-thumb scenario (conservative):
  • Targeted content leads to 5–10 booked calls/week
  • With ~4 closed deals/month, at $30k contract/month
  • Implies ~$1.5M/year revenue from content alone

Operating guidance

• Posting cadence: speaker claims twice daily across platforms; “hundreds of posts/month”
• Production method:
  • Study what goes viral in the niche
  • Remix structure into your own voice using notes/tools (mentions Claude / an “Origami prompt” workflow)
• Content pillars (YC-style):
  • posts targeted at the ideal customer
  • milestones, story, customer problems (not generic memes)

---

Step 4: Hyper-targeted warm outbound to convert the “95% iceberg”

The video distinguishes:

• Content-converted group (small)
• The larger group who sees content but doesn’t book

Framework

• “95% iceberg”: 95% of content viewers are “lurkers” (warm, not booked yet)
• Convert them using warm outbound based on observed engagement:
  • liked/commented
  • followed recently
  • visited website
  • signed up (free plan)
  • subscribed, etc.

Case study

• Goji Berry:
  • 0 → $4M annual revenue in <1 year
  • Video claims most growth came from:
    • content producing the “5%” bookers
    • layered conversion via warm outbound for the remaining 95%

Channel + tactic

• Targeted LinkedIn outbound emphasized as the “gold mine”
  • It’s social (not an inbox), so less “spam” friction
  • Credibility advantage: prospects see a real profile

Concrete targets / throughput (LinkedIn)

• Connect with ~200 new prospects/week (speaker/benchmark)
• Automation allowed (speaker mentions Origami tool, ~$29/month)
• Compliance/limits: stay within LinkedIn limits to avoid account issues
• Claimed mechanics:
  • ~50% accept rate
  • message ~100 people/week
  • “guarantee” (asserted) booking calls + deals if messaging is tailored

---

Consolidated “playbook checklist” (as presented)

• Step 1: Choose an “easy” market—rebuild legacy industries; avoid saturated “software-for-software” competition.
• Step 2: Get hundreds of customer conversations (or physically embed in workflows) until value opportunities become obvious.
• Step 3: Produce highly targeted content that the algorithm distributes to the buyer; optimize for conversions, not impressions.
• Step 4: Build warm lead lists from engagement and run hyper-targeted warm outbound, with LinkedIn as a preferred channel.

---

Presenters / sources mentioned

YC founders / companies (examples cited)

• Corgi (CEO Nico)
• Lagora
• Candid Health
• Areni
• Sim (founder Emir)
• PlanGrid (founder Tracy)
• Goji Berry

Speaker (video narrator)

• Not named in the subtitles

YC (general source)

• Y Combinator / YC batches (e.g., “current batch,” “last batch,” “P26”)