Startup Distribution Strategy: Ben Horowitz's Framework for Choosing the Right Sales Channel (2025)
Your Distribution Strategy Is a Function, Not a Preference
When Ben Horowitz — co-founder of Andreessen Horowitz and author of The Hard Thing About Hard Things — asks new entrepreneurs about their distribution model, he often gets answers like:
"I don't want to hire any of those Rolex-wearing, BMW-driving, overly aggressive enterprise sales slimeballs, so we are going to distribute our product like Dropbox did."
This answer demonstrates a deep misunderstanding of how sales channels should be designed.
The problem isn't stereotyping (though there's that too). The problem is that this founder is designing their distribution strategy around their personality instead of around their product and target market. And products often fail not because the product is bad — but because the company chose the wrong route to market.
The Distribution Formula
Horowitz presents distribution design as a mathematical function:
f(p, t) = c
p = Product
What you've built
t = Target
Who you're selling to
c = Channel
Your route to market
Critical: The function accepts only one product and one target as input. Multiple products or targets may require multiple channels.
Your channel is the output — not the starting point. You don't decide "we're going to be viral like Dropbox" and work backward. You analyze your product's delivery and assistance requirements, understand your target's decision-making process, and the right channel emerges.
Product Dimension 1: How Is It Delivered?
The first input to the distribution function is how the product reaches the customer:
A. Delivered Online
The product can be accessed, downloaded, or activated digitally without physical intervention.
Examples: Box, Dropbox, Okta, Salesforce, Workday
B. Requires Physical Delivery
The product needs a delivery service, a salesperson, or a retail store.
Examples: Smart watch, robot toy, smart security camera
Product Dimension 2: How Much Assistance Is Required?
The second dimension is the amount of help the customer needs to start using the product. This varies across a spectrum:
No Assistance Required
Sign up and start using it immediately. Self-serve all the way.
Examples: GitHub, Slack (for small teams)
Minor Assistance Required
Some setup, configuration, or guided onboarding needed.
Examples: Okta, Salesforce (for small companies)
Major Assistance & Professional Services Required
Consulting, data migration, integration, custom configuration — weeks or months of implementation.
Examples: Palantir, Pivotal, Oracle Financials; Dropbox/Salesforce/Slack for large enterprises
⚠️ The Same Product Can Require Different Assistance Levels
Dropbox for a 5-person startup? [A,A] — no assistance needed. Dropbox for a 50,000-person enterprise with data migration from legacy servers, role permissions, and security compliance? [A,C] — major professional services required. The assistance level changes based on the customer profile.
The Product Classification Matrix
Combining delivery method and assistance level creates 6 product profiles:
| Profile | Description | Examples |
|---|---|---|
| [A,A] | Online, no assistance | GitHub; Dropbox, Slack (small cos) |
| [A,B] | Online, minor assistance | Okta, Salesforce (small cos) |
| [A,C] | Online, major assistance | Oracle Financials, Palantir; Dropbox/Salesforce/Slack (large cos) |
| [B,A] | Physical, no assistance | Apple Watch, Anki Overdrive |
| [B,B] | Physical, minor assistance | Nest Thermostat |
| [B,C] | Physical, major assistance | EMC Symmetrix |
The Target Decision-Making Spectrum
The second input to the distribution function is understanding how your target makes decisions. The key insight: it's not the size of the organization that matters — it's the manner in which they decide.
Individual
A direct consumer or a single decision maker in a company
Small Group
A small team deciding together — e.g., an engineering team evaluating Trello
Entire Small Company (<1,000 employees)
Company-wide decision with clear decision makers who can act quickly — e.g., choosing Zenefits for HR
Large Group
Multiple decision makers with different economic and technical motivations
Multiple Groups Simultaneously
Different departments needing to agree — e.g., sales and marketing both deciding on marketing automation
Entire Large Company
Company-wide decision — e.g., choosing Workday for HR across 50,000 employees. Nobody inside actually knows how the decision will get made.
Matching Product to Target: The Channel Rules
Now the function executes. Here's how product classification maps to the right sales channel for each target type:
Targets I & II — Individual / Small Group
Simple decision: "Can this help me? Is it worth the money and effort?" These targets can be sold via:
- • Marketing (content, brand, paid)
- • Viral distribution (if the product inherently spreads by word of mouth)
- • Optionally telesales (especially for [A,B] products)
Death trap: If you're targeting I or II with an [A,C] or [B,C] product, "God help you as you will surely go bankrupt." The cost of providing major assistance will exceed your revenue from individual/small-team buyers.
Target III — Entire Small Company
More complex — multiple people involved, but clear decision makers who can move quickly. Generally requires a human in the loop: a salesperson who can navigate the decision-making process, explain economic and technical benefits, and set buying criteria beyond features.
Warning: [A,C] and [B,C] products are typically too expensive to sell at this level — the cost of sale outstrips the profit margin.
Targets IV-VI — Large Group / Enterprise
Complex decision-making processes where nobody inside the customer actually knows how the decision will get made. How often does a company buy an HR system? Once a decade. They won't have a process for it.
These targets require experienced, in-person sales representatives who can:
- • Navigate political dynamics across multiple stakeholders
- • Help the customer design their own buying process
- • Manage different economic and technical motivations
- • Handle security, compliance, and integration requirements
Exception: [A,A] products might not need as much human assistance initially because they can spread at the department level without requiring a complex company-wide decision.
The Dropbox vs. Box Case Study
Horowitz uses Dropbox and Box to illustrate the framework perfectly:
Dropbox
Built a brilliant [A,A] product and used primarily viral distribution. Successfully reached Targets I and II (individuals and small teams).
The gap: Had less success with Target VI (entire large companies) using that same channel. Viral adoption within departments didn't translate to enterprise-wide purchasing decisions.
Box
Saw the gap and built a direct sales channel (including those Rolex-wearing reps) to effectively reach Targets IV through VI.
The evolution: To serve enterprise targets, Box moved its product from [A,A] to [A,C] — fulfilling high-end requirements like enterprise security, admin controls, and integrations that large organizations demand.
Three Sales Channel Types
Once you've designed your channel strategy, there are three fundamental types of sales forces to staff it with:
Viral / Product-Led
Product design that enhances organic spread — forwarding, sharing, inviting. The product sells itself through usage.
Best for: [A,A] products targeting I-II
Inside Sales (Telesales)
Sales reps who work remotely — phone, email, video. Can handle minor complexity and guidance without in-person meetings.
Best for: [A,B] products targeting II-III
Direct / Field Sales
People directly in the field, on-site with customers. Can navigate complex multi-stakeholder decisions and long sales cycles.
Best for: [A,C] or [B,C] products targeting IV-VI
🚨 The Fatal Mistake
"If you insist on designing your channel based on your personality, inability to handle the diversity required to mix salespeople and engineers, or a lust for Drew Houston's business model, then please give my condolences to your employees."
— Ben Horowitz
Beyond the Basics
Once you've got the basic design, Horowitz recommends considering additional parameters:
Verticals
How does the Financial Services buying process differ from Defense & Intelligence? Different industries have fundamentally different decision-making cultures.
International
Do you sell the same way in France as in Japan? Cultural and regulatory differences often require distinct channel strategies per geography.
🚀 Find Investors Who Understand Your Distribution Model
The best investors evaluate your go-to-market as rigorously as your product. Search 100K+ verified VC and angel investor profiles on Datapile — filter by stage, sector, and investment thesis to find investors who've scaled companies with your distribution model.
Search Investors →Tagged with
Investor Outreach Template Pack
Get our proven email templates, pitch frameworks, and investor research guides — used by 1,000+ founders.
- Cold email templates that get 40%+ open rates
- Follow-up sequence frameworks
- Investor research checklist
