Growth loops explained: viral, content and paid loops for product managers

What is a growth loop?

A growth loop is a cycle in which the output of one user's action becomes the input that acquires the next user. The Reforge essay "Growth Loops are the New Funnels," co-written by Brian Balfour, Casey Winters, Kevin Kwok and Andrew Chen, defines it plainly: "Loops are closed systems where the inputs through some process generates more of an output that can be reinvested in the input." The same Reforge post says its authors had identified over 20 growth loops across acquisition, retention, defensibility and efficiency, which tells you something useful: nobody runs all of them, and most products are carried by one or two.

Here is a loop in its simplest form, using an invented product. A freelancer sends an invoice from an invoicing app. The client opens the invoice, sees the app's name at the bottom, and some share of those clients are freelancers themselves who sign up to send their own invoices. Each new freelancer sends invoices to new clients, and the cycle turns again.

The question a loop forces, in Reforge's words, is "How does one cohort of users lead to another cohort of users?"

How is a growth loop different from a funnel?

A funnel describes one pass: traffic arrives, some share signs up, some share activates, some share pays. When the period ends, you need new traffic from somewhere outside the funnel to start again. A loop describes what the users you already have produce, and how that output returns as new traffic.

The practical difference shows up in team structure. The Reforge essay points out that funnels tend to get split by stage: "Marketing owns acquisition. Product owns retention. Sales (if B2B) owns revenue." A loop cuts across all three, because the invoice that acquires the next freelancer is a product feature, a distribution channel and a revenue event at the same time. If three teams own those three things separately, nobody owns the loop.

Funnels still matter inside a loop. Every loop has conversion steps (did the client open the invoice, did they click through, did they sign up), and you improve a loop by finding the weakest of those steps. The loop is the model of how you grow; the funnel is the tool for fixing one step of it.

How does a viral loop work, and how do you measure it?

A viral loop runs when a user's normal use of the product exposes it to someone new. Invites to a shared workspace, a document shared with a non-user, a meeting link that requires joining: in each case the new person arrives because the existing user needed them there.

Casey Winters, who led growth at Pinterest and Grubhub, is sceptical of how often teams claim virality they do not have. Speaking to First Round Review about Pinterest, he said: "It wasn't clear how if I invited you, Pinterest would get better for me. Whereas for Snapchat and Facebook, that's more obvious." That is the test for viral fit. If inviting someone does not make the product better for the inviter, the invite is a favour, and favours do not compound.

Four numbers tell you whether a viral loop is running:

Metric	What it measures	How to calculate it
Invite rate	Share of active users who send at least one invite in a period	Users who invited ÷ active users
Invites per inviter	How many invites each inviting user sends	Invites sent ÷ users who invited
Invite conversion rate	Share of invites that become activated users	Activated invitees ÷ invites sent
Branch factor	New activated users produced per existing user	Invite rate × invites per inviter × invite conversion rate

A worked example with made-up numbers: if 20 percent of active users invite someone, each inviter sends 4 invites, and 25 percent of invites turn into activated users, the branch factor is 0.2 × 4 × 0.25 = 0.2. Every 100 users produce 20 more. A branch factor under 1 does not mean the loop failed. It means the loop amplifies users you get from other channels rather than sustaining growth alone, which is where most real viral loops sit.

How does a content loop work, and how do you measure it?

A content loop runs when content, created by users or by the company, gets discovered by people who are not yet users, and some of those people join and create or share more content. Winters describes it to First Round as publishing and sharing media that "triggers signups, activations and user engagement, which leads to more sharing of the media."

The two common shapes are user-generated content that search engines index (a public recipe, a public question and answer, a public profile) and company-generated content assembled from product data (a landing page for every city and category, built from listings). Winters did the second at Apartments.com and Grubhub, grouping listings into regional and category pages that search engines could find.

Measure a content loop with three numbers: how much new indexable or shareable content each cohort of users creates, what share of that content gets discovered by non-users (search sessions or social views per piece), and what share of those visitors sign up. When the loop stalls, one of the three has dropped, and it is usually the middle one: content exists but nobody outside finds it.

How does a paid acquisition loop work, and what proves it?

A paid loop closes when revenue from newly acquired users funds the ads that acquire the next users. Winters told First Round that "paid acquisition loops are very easy to get going," which is both the appeal and the trap: you can start spending in minutes, long before you know whether the spend comes back.

The metric that proves a paid loop is payback period: how long it takes the gross margin from a new customer to repay the customer acquisition cost. If a customer costs €60 to acquire and brings in €20 of gross margin a month, payback is three months, and you can recycle the same budget four times a year. If payback is eighteen months, the loop only turns if you raise money to bridge it, which is not a loop at all.

Winters is blunt about the ceiling: "Unless there is a core network effect inside of a product, paid acquisition is a race to the bottom over time." The first users you buy convert best. As spend scales, you reach people who fit less well, cost more, and retain worse, so a paid loop needs a payback period short enough to survive that decline.

How do you pick the right loop for your product?

Pick the loop that matches what users already do, not the one you admire in someone else's product. The signals are usually visible before you build anything:

Loop	Fit signal in current usage	Proof metric	Typical failure
Viral	Users already need other people inside the product to get value	Branch factor, invite conversion rate	Invites exist but add nothing for the inviter
Content	Using the product naturally produces something public and searchable	Content created per user, discovery rate, visitor sign-up rate	Content is created but never found by non-users
Paid	Users pay early, and margin per user is high enough to fund acquisition	Payback period against a target you set in advance	Early cohorts pay back, later ones never do

The marketplace version of this choice deserves its own note, because the loop crosses sides. Winters describes running paid campaigns on top of Grubhub's content loop: "Any new user who ordered food made it more likely that restaurants would join the site, and more restaurants meant more users." On a marketplace, a loop is judged by whether each side's growth makes the other side more valuable, which is why marketplace liquidity sits underneath any marketplace loop.

Are growth loops just funnels with an arrow drawn back to the start?

The fair objection is that a loop diagram can make anything look self-sustaining. Draw an arrow from "happy customer" back to "new visitor" and you have a loop on a whiteboard, whether or not a single happy customer ever brought in anyone.

The objection lands whenever the returning arrow has no number on it. A loop earns the name only when you can measure the step where output becomes input: invites sent and accepted, pages indexed and visited, revenue recycled into spend. If you cannot put a metric on that step, you have a funnel with good intentions.

The second weakness is that every loop assumes retention. Andrew Chen, citing Quettra data, reported that "the average app loses 77% of its DAUs within the first 3 days after the install." That figure comes from Android app data collected in 2015 and says nothing about your product specifically, but the direction holds everywhere: users who leave in week one do not invite, create content or pay back their acquisition cost. Fix retention before you invest in any loop, and read what a growth product manager owns for why that order matters.

Where does a product manager learn to build and measure growth loops?

The viral acquisition loop and content acquisition loop glossary entries cover each loop in isolation. For the full system, Builders Camp's Growth for Product Managers bootcamp runs 2 weeks with 4 live sessions and 8 self-paced microlessons, directed by Andre Albuquerque, and is part of the Product Management Starter Track and the Growth Specialist Track. Its published syllabus includes comparing viral, content and paid acquisition loops and validating which fits your product and how to instrument it, alongside retention, monetization and cohort analysis.

Before you build any loop, write down the single metric that would prove it is turning, and the number that would make you stop. A loop without that number is a diagram.