Quick answer
A growth loop is a closed system in which an input passes through a series of actions and produces an output that can be reinvested as an input to a future cycle. Examples include users inviting collaborators, customers creating indexable content, or revenue funding efficient acquisition. A credible loop specifies the unit, conversion at every step, cycle time, retention, quality, capacity and saturation. It should be modeled by cohort and validated against observed behavior. Funnels remain useful for diagnosing a journey segment, while loops explain how one cohort can help create or reactivate another.
What is a growth loop?
A growth loop describes how a product or business turns an input into an output that can generate another input. Reforge popularized loops as a system-level alternative to explaining company growth only through a one-direction funnel. The central question is how one cohort contributes to a future cohort.
A circle on a slide is not sufficient. The mechanism must name actors, actions, objects, distribution and reentry. An invitation that is never accepted, a page no one discovers or revenue that cannot be reinvested economically is an output, but it has not completed a productive loop.
Why loops complement funnels
Funnels show conversion and loss through a bounded journey. They are excellent for locating an activation drop or a checkout failure. Their usual form does not show how what exits one journey may replenish the beginning, so teams can default to buying or pushing more top-of-funnel volume.
Loops add that feedback structure. They connect product behavior, distribution and monetization in one model. A team can still use a funnel inside each loop step, but it evaluates whether customer value, retention and reinvestment create repeatable future input rather than a temporary spike.
The anatomy of a growth loop
Define the starting unit and one complete cycle. Map the value-producing action, reusable output, distribution channel and condition for qualified reentry. Draw leakage explicitly. Every arrow should correspond to an observable conversion or delay rather than an assumed handoff.
Then add stocks and constraints: retained active users, content inventory, marketplace supply, budget, review capacity and channel reach. A loop may depend on an external platform whose ranking or policy changes. Make that dependency visible instead of treating distribution as automatic.
Input
Name the person, object or resource that enters one cycle.
- What exactly enters?
- Is it new, returning or paid?
Value
Describe the customer action that creates genuine utility.
- Why would a user act?
- What proves value?
Output
Identify the reusable result created by the action.
- What is produced?
- Does quality vary?
Distribution
Explain how the output reaches an eligible future participant.
- Which channel carries it?
- Who controls that channel?
Reentry
Measure whether the output becomes another qualified input.
- What completes the cycle?
- How long does it take?
Common types of growth loop
Viral and collaboration loops create invitations or shared artifacts that attract other users. User-generated content loops produce pages, listings or media that a company or community distributes. Marketplace loops add supply that improves selection or liquidity and can attract demand, which may in turn attract more supply.
Paid loops reinvest contribution into acquisition, while sales loops use successful customers, proof and expansion to create future qualified demand. Retention loops return existing users through ongoing value, reminders or fresh content. A business can have several loops, but one or two often deserve primary operating attention.
How to map and build a growth loop
Start with interviews, event paths, channel evidence and economics. Ask how the strongest retained cohort first arrived, reached value, created an output and influenced someone else or itself later. Build a qualitative map before a spreadsheet so false precision does not conceal an implausible mechanism.
Instrument the weakest unknowns, estimate ranges and compare the model with historical cohorts. Prioritize a step that can improve customer value and whole-cycle output. Test it, observe delayed effects and revise the map. Treat the model as versioned operational knowledge, not a permanent truth.
- Starting unit explicit
- One cycle boundary defined
- Value action observable
- Output reusable
- Distribution mechanism named
- Qualified reentry measured
- Leakage shown
- Cycle time estimated
- Retention included
- Quality threshold documented
- External dependencies recorded
- Saturation and guardrails monitored
Growth loop example
Fieldnote's hypothetical content loop begins with useful research, not with indexed pages. Publication produces a distributable artifact, but the loop completes only when a qualified reader receives value, activates and eventually contributes another quality artifact.
This distinction prevents teams from flooding search with thin pages. The quality threshold, user publication choice and downstream activation are part of the mechanism. If any of them erodes, raw page growth can coexist with a weakening loop.
Fieldnote is a hypothetical collaborative notebook for research teams. Some users publish selected notes on the open web, where relevant readers can discover the work. The team wants to know whether this behavior forms a durable content loop rather than merely generating page views.
An activated researcher creates a useful note, chooses to publish it, and adds enough context for an outside reader. Search or direct sharing distributes the page. A qualified reader may join, create a project and later publish another useful note.
The model tracks eligible researchers, quality-approved public notes per researcher, indexation delay, qualified visits per note, reader activation and the share of activated readers who later create a public note.
Private work is excluded by default. Publication requires clear control, sensitive-content checks and removal options. Thin or duplicated pages are not counted as useful outputs, even if they increase the number of URLs.
Fieldnote can test better publication guidance or reader-to-project continuity at a specific weak step. It measures the whole-cycle downstream effect rather than declaring success from a rise in publishing alone.
The team compares cohorts, cycle times and topic saturation. If quality or activation declines as output grows, it revises the mechanism instead of assuming that more content always strengthens the loop.
Fieldnote and every described result are hypothetical. Actual loop models need observed cohort data, channel constraints and customer safeguards.
Quantify loop power and speed
Model each step as a rate or distribution attached to a cohort. A simple invitational loop might use active inviters per cohort, invitations per inviter, delivery, acceptance, activation and retained contribution. Multiplying averages can offer a rough scenario, but distributions, clustering and repeated invitations complicate inference.
Cycle time matters alongside output. A loop that returns one qualified input after a day behaves differently from one that returns it after a year. Use sensitivity analysis to find which assumptions control the result, and compare forecasts with observed cohorts before allocating major resources.
Operate teams around the loop constraint
A loop crosses functional boundaries. Search content may require product, editorial, engineering, legal and SEO; collaboration loops may require activation, invitation delivery and team administration. Give a cross-functional owner the mandate to improve the whole mechanism while domain owners retain quality standards.
Review input, step conversion, time, quality and downstream value together. When one constraint improves, another may become limiting. Update goals accordingly and avoid keeping a team attached to an input metric after the bottleneck has shifted.
Govern loop quality and external effects
Compounding systems can compound harm. Invitation loops can become spam, content loops can amplify misinformation, marketplace loops can reduce participant welfare, and paid loops can hide poor contribution economics. Define eligibility, quality and abuse controls before optimizing output volume.
Respect privacy, consent, intellectual property, accessibility and removal rights. Monitor complaints, moderation load, channel policy and concentration. A sustainable loop creates value for the participant entering each cycle, not only for the company measuring growth.
Limitations and common loop mistakes
A loop diagram can imply certainty where behavior is noisy and context-dependent. External distribution may saturate, competitors may copy features, retained users may tire, and outputs can cannibalize one another. Growth can also come from brand, seasonality or one-time events outside the mapped loop.
Common mistakes include calling any circular journey a loop, omitting retention, multiplying unvalidated averages, optimizing one step, ignoring cycle time and counting low-quality outputs. Use loops to express a testable growth hypothesis. Keep funnels, cohort analysis, causal evaluation and financial models alongside them.
A growth loop earns its name when a measured, valuable output reliably creates another qualified input.
Frequently asked questions
What is the difference between a growth loop and a funnel?
A funnel describes progression and loss through a journey. A loop explains how an output from one cycle can replenish an input to another. A funnel can model steps inside a loop.
What is a viral growth loop?
It is a loop in which users create invitations, shared experiences or artifacts that attract other eligible users who can later repeat the behavior.
How do you measure a growth loop?
Track conversion, delay, retention and quality at every step by cohort, plus the rate at which outputs become qualified reentries. Validate a quantitative model against actual behavior.
Can a business have multiple growth loops?
Yes. Content, collaboration, paid, sales and retention loops can coexist. Identify the primary mechanisms and their dependencies instead of drawing every possible circle.
When is a loop not sustainable?
When value, retention or economics are weak; output quality declines; distribution depends on a saturating channel; or growth imposes unacceptable customer, social or operational costs.
Sources and further reading
- Reforge: Growth Loops Are the New Funnels ↗Primary practitioner articulation of closed growth systems and reinvested outputs
- Reforge: The Racecar Growth Framework ↗Distinguishes compounding engines from one-time boosts, efficiency improvements and fuel
- Reforge: Product-Led Growth Is a Major Growth Motion ↗Explanation of product growth loops across activation, engagement and monetization
- ACM: The Dynamics of Viral Marketing ↗Primary empirical research on person-to-person recommendation dynamics and network structure