Purpose

This recipe produces a formal classification of a startup idea into one of 12 primary business model types, with secondary type identification for hybrid models. The output determines which downstream playbooks, metrics, capital requirements, and go-to-market strategies apply. Misclassification wastes months pursuing wrong benchmarks and investor expectations.

Prerequisites

Startup idea description (minimum 2-3 sentences describing what the product does and for whom)
Revenue hypothesis (even rough — "we'd charge monthly" or "we'd take a cut of transactions")
Target customer segment (B2B, B2C, B2B2C, or B2G)

Constraints

Most startups blend 2-3 models. Classify by primary revenue driver, not feature list. A SaaS tool with a marketplace feature is still SaaS if subscriptions drive >60% of revenue. [src1]
Hardware startups require 3-10x more capital than pure software. This classification directly affects fundraising strategy and investor targeting. [src6]
AI-native startups should be classified by delivery model (API service, embedded AI, platform), not by the technology itself. [src3]
Classification is not permanent. Many startups pivot model types. Revisit classification after each major pivot. [src4]

Tool Selection Decision

Path	Tools	Cost	Speed	Output Quality
A: Direct Classification	Text editor	$0	15 min	High
B: Revenue Discovery	Questionnaire + text editor	$0	30 min	Medium-high
C: Batch Classification	Spreadsheet	$0	45 min	High
D: Automated Extraction	LLM + structured output	$0.01-0.05	5 min	Medium

Execution Flow

Step 1: Gather Idea Signals

Duration: 10 minutes. Collect 8 signals covering delivery method, revenue model, customer segment, supply side, network effects, marginal cost, switching cost, and moat type.

Step 2: Apply Classification Decision Tree

Duration: 10 minutes. Map signals through the decision tree to arrive at primary business model type from 12 categories: SaaS, Marketplace, Hardware/D2C, E-Commerce, Professional Services, Productized Service, Tech-Enabled Services, API/Infrastructure, Developer SaaS, Media/Content, Network/Social, Subscription Media.

Step 3: Score Hybrid Model Fit

Duration: 10 minutes. For multi-type ideas, score each candidate on 5 weighted dimensions: revenue alignment (0.30), operational alignment (0.25), investor expectation (0.20), scaling path (0.15), competitive positioning (0.10).

Step 4: Map to Applicable Playbooks and Metrics

Duration: 10 minutes. Assign primary metric, key playbooks, capital requirement range, and timeline to revenue based on classification type.

Step 5: Generate Classification Report

Duration: 5 minutes. Compile JSON report with primary type, secondary type, confidence score, applicable playbooks, and classification rationale.

Quality Benchmarks

Quality Metric	Minimum	Good	Excellent
Signal completeness	> 6/8	8/8	8/8 with notes
Classification confidence gap	> 0.5	> 1.0	> 2.0
Playbook mapping	≥ 3	≥ 5	≥ 7 with sequencing

Error Handling

Error	Cause	Recovery
Idea maps to 3+ types equally	Idea too broad	Define initial wedge product, classify that
No matching type in tree	Novel business model	Classify by closest analog, document deviations
Founder disagrees	Different mental model	Review signals together, re-score

Cost Breakdown

Component	Free	Paid	At Scale
Classification framework	$0	$0	$0
Research validation	$0	$50-100	$500+
Facilitation	$0	$200-500	$2,000+

Anti-Patterns

Wrong: Classifying by technology instead of business model

Calling it an "AI startup" tells nothing about go-to-market, metrics, or capital needs. [src1]

Correct: Classify by delivery and revenue model

Use the 8-signal framework to determine how value is delivered and how money flows.

When This Matters

Use this recipe as the first step in any startup planning pipeline. Every downstream decision depends on correct business model classification.

Startup Idea Classification Framework