GrowthBooks

GrowthBooks — Evidence & Evaluation Definitions

Canonical definitions used across all GrowthBooks evaluators

Version: v1.0

Status: Canonical · Locked

Applies to:

  • • Demand Reality (Top of Funnel)
  • • Conversion Clarity (Mid Funnel)
  • • Revenue Engine (Bottom of Funnel)

Purpose of This Document

This document exists to prevent evaluator drift, protect scoring integrity, and ensure consistent application of evidence standards across all GrowthBooks evaluations.

  • Prevents inconsistent interpretation of GTM signals
  • Makes explicit what is scored and what is ignored
  • Establishes a shared reference for disputes or clarification
  • Does not provide tactics, advice, or recommendations

If information does not meet the definitions below, it scores zero.

How to Use This Document

  • Read before running any GrowthBooks evaluator
  • Refer to when unsure whether something qualifies as evidence
  • Treat this as a classification system, not guidance
  • Definitions describe what *is*, not what *should be done*

Core Evidence Concepts (All GrowthBooks Evaluators)

Evidence

Evidence is externally verifiable information that demonstrates observable buyer or market behaviour.

Qualifies

  • Buyer-initiated inbound leads
  • Unsolicited buyer conversations
  • CRM pipeline records
  • Deal progression data
  • Cleared payments or contracts

Does not qualify

  • Opinions or confidence
  • Internal plans or strategy
  • Forecasts without history
  • Slides, narratives, or intent

Rule: If it cannot be shown to an evaluator, it is not evidence.

Behaviour vs Opinion

Behaviour is what buyers do. Opinion is what they say.

Qualifies

  • Buyer initiates contact
  • Buyer progresses to next step
  • Buyer accepts pricing
  • Buyer pays

Does not qualify

  • "This sounds interesting"
  • "We would use this"
  • Praise or encouragement

Rule: Behaviour scores. Opinion does not.

Signal vs Metric

A metric is a number. A signal is a number that reflects buyer choice.

Qualifies as signal

  • 5 unsolicited inbound leads in 30 days
  • 70% progression to next step
  • Deals closing at list price

Does not qualify

  • Website traffic
  • Impressions
  • Followers
  • Activity volume without buyer action

Rule: Metrics only matter when they represent decisions.

Founder Influence

Founder influence exists when demand, conversion, or revenue depends on founder identity, reputation, or direct involvement.

Qualifies as founder-influenced

  • Deals sourced from personal network
  • Inbound driven by founder visibility
  • Founder required to close

Rule: Founder dependence is penalised, even if outcomes look positive.

Missing Data

Information that is unknown, unavailable, or not tracked.

Rule: Unknown data scores zero. No inference is allowed.

Evidence Weighting (All Evaluators)

Highest Weight

  • Buyer payments
  • Signed contracts
  • Irreversible commitments

High Weight

  • Buyer-initiated inbound
  • Deal progression

Medium Weight

  • Buyer choice under constraint

Low Weight

  • Buyer statements

Zero Weight

  • Assumptions
  • Hope
  • Vision
  • Effort

Evidence by GrowthBooks Evaluator

Demand Reality — Top of Funnel

Evaluates whether real, observable market demand exists before outbound effort or persuasion by assessing ICP self-recognition, urgency, and unsolicited interest.

What it evaluates

  • Whether a specific ICP recognises itself
  • Whether the problem is urgent and triggered
  • Whether interest appears without prompting
  • Whether demand exists beyond founder networks

Qualifying evidence

  • Buyer-initiated inbound leads
  • Unsolicited buyer conversations
  • Repeat inbound from the same ICP
  • Unprompted referrals

Does not qualify

  • Outbound responses
  • Warm introductions
  • Founder-sourced conversations
  • "People seem interested"

If demand disappears when effort stops, it is not demand.

When to use

When pipeline exists but intent, urgency, or inbound pull is unclear.

Run free evaluator

Conversion Clarity — Mid Funnel

Assesses whether interested prospects clearly understand the value, outcome, and next step required to convert — without sales intervention or explanation.

What it evaluates

  • Buyer understanding of value
  • Message consistency across touchpoints
  • Proof strength and credibility
  • Structural friction or stalls

Qualifying evidence

  • Prospects describe the value correctly
  • Low frequency of repeated explanation
  • Proof matches buyer risk
  • Clear progression rates

Does not qualify

  • "Sales can explain it"
  • Repeated clarification
  • Weak or mismatched proof

If explanation is required, clarity does not exist.

When to use

When traffic or leads exist, but conversion depends on persuasion or founder explanation.

Run evaluator

Revenue Engine — Bottom of Funnel

Evaluates whether revenue generation is repeatable, accepted by the market, and structurally independent of founder effort.

What it evaluates

  • Pricing acceptance
  • Sales efficiency
  • Repeatability of motion
  • Founder independence

Qualifying evidence

  • Deals closing at stated price
  • Predictable sales cycles
  • Repeatable process across deals
  • Deals closing without founder

Does not qualify

  • Heavy discounting
  • Bespoke deals
  • Founder-required closes

Repeatability beats volume. Founder dependence is penalised.

When to use

When growth feels busy, but revenue predictability or ownership is fragile.

Run evaluator

Common Misinterpretations (Not Scored)

Passion
Hustle
Intelligence
Vision
Slide quality
"We're early"
"Everyone liked it"

Scoring Implications

  • Missing evidence = zero
  • Founder dependence = penalty
  • Repeated confusion = penalty
  • Misaligned proof = zero
  • Strong claims without evidence are penalised

The evaluator does not negotiate. The rubric is fixed.

Final Canonical Statement

This document defines how go-to-market reality is assessed inside GrowthBooks.

It does not define what to build, change, or believe.

Comfort indicates insufficient rigor.