Distribution

How your product reaches customers. Peter Thiel’s most underappreciated insight from Zero to One: “Poor sales rather than bad product is the most common cause of failure.” Most founders obsess over building; the winners obsess over distributing.

The Power Law of Distribution

Distribution follows a power law: one channel will likely be far more powerful than all others combined for any given business. If you nail one channel, you have a great business. If you try several but don’t nail any, you’re finished.

The implication: don’t diversify distribution early. Find the one channel that works and pour everything into it.

Thiel’s Distribution Spectrum

Thiel maps distribution channels by customer lifetime value (CLV) and customer acquisition cost (CAC):

ChannelCLV RangeCAC RangeHow It WorksExample
Complex sales$1M-$100M+$100K+Personal relationships, months/years per dealSpaceX selling to NASA
Enterprise sales$10K-$1M$1K-$100KSales teams, demos, pilotsPalantir, Salesforce
The Dead Zone$1K-$10K???Too expensive for ads, too cheap for sales teamsMost B2B SaaS fails here
Marketing & ads$1-$1K$1-$100Mass reach, brand, SEO/SEMConsumer packaged goods
Viral distribution$0-$100~$0Product functionality drives sharingPayPal, Facebook, WhatsApp

The Dead Zone

The most dangerous region: products priced between ~$1K-$10K. Too expensive for self-serve/advertising, too cheap to justify dedicated sales reps. Many startups die here because no distribution channel is economically viable.

Solutions:

  • Move upmarket (raise prices to justify sales)
  • Move downmarket (reduce prices to enable self-serve)
  • Find a creative channel that bridges the gap (Collison installation)

Viral Distribution

The most powerful channel when it works. A product is viral when its core functionality encourages users to invite others:

  • PayPal: Every payment invites the recipient to create an account
  • Slack: Every workspace invites colleagues
  • Dropbox: Every shared folder invites the recipient
  • WhatsApp: Every message requires the other person to have the app

Viral coefficient > 1.0 means exponential growth. Even 0.5-0.9 significantly reduces CAC.

The Fundamental Equation

CLV must exceed CAC. The higher your price, the more you can spend to acquire a customer — and the more it makes sense to invest in sales.

If CLV is…Distribution approach
<$100Must be viral or mass marketing. Can’t afford salespeople.
$100-$1KSelf-serve with content/SEO. Light-touch sales at most.
$1K-$10KDead zone. Need creative solutions.
$10K-$100KInside sales teams, demos, free trials → paid.
$100K+Field sales, relationship-driven, months-long cycles.

Distribution for Startups (Practical)

Altman’s hierarchy from the Startup Playbook:

  1. Word of mouth from product-market-fit — the best channel, can’t be bought
  2. Manual recruitmentdo-things-that-dont-scale in early days
  3. Referral programs — incentivize existing users to invite others
  4. Paid acquisition (SEO/SEM, ads) — recover CAC within 3 months
  5. Direct sales — viable for >$500 LTV products
  6. Partnerships — almost never work early (PG, Altman both warn)

The First 1,000 Users (Lenny Rachitsky)

Data from 40+ consumer apps: just 7 strategies account for all early growth. Most apps used exactly one. None used more than three.

StrategyHow It WorksFamous Example
1. Go offlineFind users physicallyTinder → sororities at USC
2. Go onlineFind users in digital communitiesDropbox → Hacker News video
3. Invite friendsPersonal network seedingFacebook → dorm list → 1,200 in 24hrs
4. Create FOMOWaitlists, invite-onlyRobinhood → 600K waitlist
5. Leverage influencersKey people amplify youTwitter → Om Malik blog post
6. Get pressMedia coverage drives signupsSlack → 8K requests on launch day
7. Pre-launch communityBuild audience before productProduct Hunt → email list of 30

The pattern: every strategy is doing things that don’t scale. Direct engagement, not paid channels.

The rule: Pick ONE strategy and nail it. Don’t spread across all seven.

Selling to Developers Is Different

Adam Frankl (first VP Marketing at 12 dev tool startups, 3 unicorns: JFrog, Neo4j, Sourcegraph) codified the hardest distribution challenge: developers don’t respond to traditional marketing. The central principle:

“Developers don’t lie to other developers.”

This single idea reshapes everything:

  • Peer validation is the dominant signal
  • Polished marketing feels like lying
  • Traditional funnels don’t work — developers research independently and non-linearly
  • You can’t fake credibility; developers will find out

The Developer Adoption Pattern

The only path that works:

  1. Individual developer tries the tool (low friction = critical)
  2. Developer tells their team
  3. Team adopts informally
  4. Team lead champions it to engineering leadership
  5. Engineering leadership signs enterprise contract

Top-down sales doesn’t work until step 5. Trying to force it earlier destroys trust.

Tactics That Work with Developers

WorksDoesn’t Work
Founder posts daily on social media (real voice)Polished press releases
Sell the category first (“why graph databases matter”)Sell the product first
Technical Advisory Board of real usersCustomer Advisory Board (too business-focused)
Open source + paid enterprise tierEnterprise-only with a demo request
Detailed docs, changelogs, GitHub issuesGlossy landing pages
”Talk to large numbers of developers""Be clever in a conference room”
Peer recommendations (HN, Reddit, coworkers)Paid ads

Sell the Category, Not the Product

Stripe had to teach developers that online payments could be simple. Snyk had to teach that dependencies are a security risk. Sourcegraph had to teach that code search is worth paying for. Educate developers about why this TYPE of tool matters before positioning your specific product.

Once the category exists in the developer’s head, you position within it. This is why the best dev tool companies publish deep technical content about the problem space — not about their product.

Programmatic SEO: The Scalable Channel

Most SEO is manual — writing articles one at a time. Programmatic SEO is different: it generates thousands of pages from a template + database, each targeting a specific long-tail query.

The Formula

Template + Database + Automation = Pages at Scale

Famous Implementations

CompanyPatternScale
Zapier”[App A] integration with [App B]“70K+ pages, 6.3M visits/month
TripAdvisor”[Category] in [Location]” with dietary/price/occasion attributes226M visits/month
Yelp”[Business type] near [Location]“Millions of pages
Indeed”[Job title] jobs in [City]“Millions
WebflowCMS Collections → template-rendered pagesHundreds of targeted pages

When Programmatic SEO Works

  • Genuine long-tail demand exists for the pattern (people actually search this)
  • You have real data to fill the pages (not invented content)
  • Template quality — each page serves real intent, not spam
  • Technical execution — fast pages, clean schema, proper sitemap

When It Fails

  • The search demand isn’t there — you’re building for ghosts
  • The content feels auto-generated (Google penalizes at scale)
  • Templates produce duplicate-feeling pages
  • You’re scaling volume without proportional value

The Paradox

Most teams fear creating thousands of pages because it sounds like spam. But search engines now reward scale when the content genuinely serves users. The danger isn’t volume — it’s volume without value.

The key test: would a human find each individual page useful? If yes, you can have millions of them and Google will love it.

Growth Loops: Distribution as a System

Balfour’s growth-loops framework reframes distribution from a one-time channel choice to a compounding system. Every growth loop has a distribution mechanism, and the most powerful loops embed distribution into the product itself. Product-Channel Fit — the principle that “products are built to fit with channels, channels do not mold to products” — means your product must be architecturally designed for its primary loop type. When product-channel fit breaks (as when Facebook killed the APIs Pinterest relied on), companies must rebuild for a new channel or die. See growth-loops for the full Four Fits framework and loop taxonomy.

Distribution as Moat

Once you own a distribution channel, it becomes a competitive moat:

  • Stripe owned developer distribution (API ecosystem)
  • Airbnb owned supply-side distribution (host acquisition + photography)
  • Salesforce owned enterprise sales distribution (pioneered SaaS sales model)

Competitors who build better products still lose if they can’t match your distribution.

See Also

Sources