There are seven growth loop archetypes every growth engineer should know: viral/invite, content/SEO, paid/reinvestment, sales/expansion, UGC/marketplace, data-network, and ecosystem/integration. The first four are Reforge's canonical acquisition loop taxonomy. The last three have emerged as distinct archetypes since 2023 as platform, marketplace, and AI businesses scaled. This article maps each one with two examples, the failure mode, and the single leading indicator that tells you whether the loop is compounding or breaking.
What are growth loops, and why do they replace funnels?
A growth loop is a closed system where the output of one cycle becomes the input of the next. Funnels are linear (in at the top, out at the bottom). Loops compound (each user, page, or dollar produces the next).
Brian Balfour, Reforge's CEO, framed it in 2018: "The fastest growing products today are better represented as a system of loops, not funnels."
The shift matters because:
- Funnels leak. Every stage loses users with no reinvestment.
- Loops compound. Each cycle reinvests the output back as input.
- Loops are defensible. A loop is harder to copy than a campaign.
Reforge originally identified four acquisition loop categories: viral, content, paid, and sales. Since then, three more archetypes have stabilised in the literature: UGC/marketplace, data-network, and ecosystem/integration. Together they form the seven-loop taxonomy used in this article.
What are the 7 growth loop archetypes in 2026?
The seven growth loop archetypes are: viral/invite, content/SEO, paid/reinvestment, sales/expansion, UGC/marketplace, data-network, and ecosystem/integration.
They differ on what compounds (users, content, dollars, data, integrations) and how fast the loop closes. Here is the full comparison:
| # | Archetype | What compounds | Loop closure speed |
|---|---|---|---|
| 1 | Viral / Invite | Users invite users | Days to weeks |
| 2 | Content / SEO | Pages produce pages | 6-12 months |
| 3 | Paid / Reinvestment | Dollars buy dollars | Days, gated by payback |
| 4 | Sales / Expansion | Revenue funds reps | Renewal cycle (12-18 mo) |
| 5 | UGC / Marketplace | Supply attracts demand | Weeks to months per geo |
| 6 | Data-Network | Usage improves product | Quarters to years |
| 7 | Ecosystem / Integration | Partners build moat | Years |
The first four are Reforge's original taxonomy. The last three are extensions that the growth community has codified as platform businesses, AI products, and developer ecosystems became mainstream.
1. What is a viral / invite growth loop?
A viral growth loop compounds when each active user invites or exposes new users from inside the product. The loop closes user-to-user, with no marketing spend in between.
How it compounds: User signs up → uses product → invites/shares → invitee signs up → repeats. Math: K-factor = invites per user x conversion rate. K > 1 = exponential.
Examples:
- Figma: every shared design file is a viral surface. Designers invite developers, developers invite PMs. Cross-functional adoption inside companies is Figma's core loop, as Kevin Kwok detailed in 'Why Figma Wins'.
- Dropbox: the classic referral viral loop -- give 500MB to inviter and invitee. Dropbox's referral program reached a K-factor of 0.35-0.40, below 1 but enough to materially drive early growth.
Failure mode: K-factor stuck below 0.3 for 90+ days. Below this threshold, paid acquisition has to fund the loop in perpetuity, which defeats the point.
Leading indicator: invites sent per active user per week. Conversion rate of those invites is downstream and lagging.
2. How does a content / SEO growth loop work?
A content growth loop compounds when content (editorial, programmatic, or user-generated) ranks in search, attracts users, and either those users or your system produce more content that ranks.
How it compounds: Page is published → ranks for query → attracts user → user signs up or generates more indexable content → more pages rank.
Examples:
- Pinterest: users save pins, pin pages get indexed, search traffic drives sign-ups, new users save more pins. Pinterest's growth is largely an SEO + UGC loop.
- Zillow: programmatic content loop. Zillow has nearly 5.2 million pages, mostly programmatically generated, driving 33M visits/month. Each property becomes a page; each page is an entry point.
This is the Rich Barton Playbook -- bootstrap a Data Content Loop to dominate search in your industry, as Kevin Kwok wrote in 'Making Uncommon Knowledge Common'.
Failure mode: Google core update deindexes thin pages. Content loops built on programmatic SEO without first-party data are existentially exposed to algorithm changes.
Leading indicator: indexed pages with Top-10 rankings, tracked weekly. Page count alone is vanity.
3. What is a paid / reinvestment growth loop?
A paid growth loop compounds when revenue from paid-acquired customers is reinvested into more paid acquisition. The loop closes through cash, not behaviour.
How it compounds: Spend $X on ads → acquire user → user pays back X + margin → reinvest into ads. Math gates everything: payback period < channel CAC inflation rate.
Examples:
- HelloFresh: meal-kit subscription model where Q1 cohort revenue funds Q2 acquisition. The loop is sensitive to churn -- churn breaks the math.
- Wish (early years): bought traffic via Facebook ads, monetized via low-cost product margin. Worked until iOS 14 ATT changes broke the targeting layer in 2021.
Failure mode: CAC payback exceeds 12 months, or a single channel exceeds 60% of spend. When Meta CPMs jumped 61% YoY in 2021-2022, dozens of paid-loop DTC brands collapsed.
Leading indicator: LTV:CAC ratio and payback period, tracked by cohort and channel. The 3:1 LTV:CAC threshold is the loop's lifeline.
4. How does a sales / expansion growth loop compound?
A sales/expansion loop compounds when closed revenue funds more reps, and existing accounts expand seat count, modules, or contract value over time. This is the dominant B2B SaaS loop for ACVs above $10k.
How it compounds: Rep closes deal → seat lands inside account → champion advocates internally → expansion seats → revenue funds another rep.
Examples:
- Salesforce: the textbook land-and-expand. Land a seat in one team, expand to the org, cross-sell into Service Cloud, Marketing Cloud, Slack.
- HubSpot: lands with free CRM, expands via Marketing/Sales/Service hubs. NRR consistently above 100% for years.
In 2026, most competitive B2B SaaS run hybrid GTM motions that combine PLG acquisition with SLG expansion.
Failure mode: Net Revenue Retention drops below 100%. When NRR < 100%, the expansion loop is leaking faster than it compounds. The company is running on net new logos, not loop math.
Leading indicator: Net Revenue Retention (NRR) by cohort, tracked monthly. Above 110% = healthy. Above 120% = best-in-class.
5. What is a UGC / marketplace growth loop?
A UGC/marketplace growth loop compounds when users create supply (listings, reviews, content), which attracts demand, which brings more users who create more supply. This is the cross-side network effect made operational.
How it compounds: Host lists property → guest searches and books → review is published → review attracts next guest + signals to next host. Each side feeds the other.
Examples:
- Airbnb: hosts list, guests book, reviews validate, photos build trust. Higher host density in a city increases guest demand, which incentivizes more hosts.
- Pinterest: users pin content, pins get indexed and ranked, search traffic finds the pins, new users sign up to pin. Both UGC and content loops fused.
Failure mode: liquidity gap. If supply or demand can't reach a critical density per geo/category, the loop never closes. Most marketplace failures die at the cold-start problem.
Leading indicator: supply/demand match rate per geographic or category cohort -- not aggregate. Aggregate liquidity hides local cold-starts.
6. What is a data-network growth loop?
A data-network growth loop compounds when usage produces data, that data improves the product (better recommendations, routing, search relevance, or AI model quality), and the improved product attracts more usage that generates more data.
This is distinct from classic network effects. With network effects, more users = more value. With data-network loops, more usage = more value, even at constant user count.
How it compounds: User uses product → generates interaction data → data trains/improves model → model produces better output → more user time spent → more data.
Examples:
- Google Search: every query, click, and dwell-time signal trains relevance. The more people search, the better the algorithm, the more they search.
- Waze: GPS traces from drivers improve routing. More drivers = more traces = better ETAs = more drivers. Matt Turck called this the canonical 'data network effect'.
In 2026, AI products like Cursor, Perplexity, and Glean are building data-network loops on user interaction data.
Failure mode: cold-start. Without enough seed data, the model output is worse than a heuristic, no one uses it, no data is generated. Most AI startups die here.
Leading indicator: model accuracy or relevance lift per 10x of training data. If 10x data only produces 5% lift, the loop has flattened.
7. What is an ecosystem / integration growth loop?
An ecosystem/integration growth loop compounds when third-party developers, partners, or integration builders create on top of your platform, which raises switching costs, increases retention, and attracts more partners because the platform is more valuable.
How it compounds: Customer adopts platform → integrates with 3+ tools or installs partner apps → switching cost rises → retention compounds → larger install base attracts more developers/partners → more integrations available.
Examples:
- Salesforce AppExchange: 9,000+ apps. Each integration deepens the lock-in for existing customers and signals to net-new customers that Salesforce is the safe enterprise choice.
- Shopify: app store + theme ecosystem + Shopify Partners program. Merchants who install 5+ apps churn at materially lower rates than those who install zero.
Customers who use integrations churn at significantly lower rates -- integration-enabled retention is the most important ecosystem metric.
Failure mode: partner economics break. If partners can'tmake money on your platform, they leave. Twitter killed its developer ecosystem twice (2012, 2023) and the loop reversed each time.
Leading indicator: active integrations per customer. Customers with 3+ active integrations have ~2-3x higher retention than zero-integration customers across most B2B SaaS benchmarks.
Which growth loop archetype should a B2B SaaS pick first?
Most B2B SaaS should start with either a sales/expansion loop or a content/SEO loop -- not a viral one.
Use this decision tree:
- ACV > $10k and natural seat expansion? → Start with sales/expansion loop.
- Buyers research solutions via search? → Start with content/SEO loop.
- Product has built-in collaboration (file shares, comments, multi-user docs)? → Viral/invite is viable as the first loop. (Figma, Loom, Notion fit this.)
- Below $1k ACV with self-serve sign-up? → Paid/reinvestment can work, gated on LTV:CAC math.
- Building an AI product? → You will need a data-network loop, but it is a second loop. Pair with content/SEO or sales/expansion to bootstrap usage first.
Avoid as your first loop:
- Marketplace (cold-start risk too high).
- Ecosystem (requires installed base of 1,000+ customers to attract partners).
The pattern across every $100M+ B2B SaaS is: nail one loop until it compounds, then layer a second. Running 3+ loops simultaneously almost always fails because each demands different team skills, metrics, and capital allocation.
For a deeper dive, see how to build your first growth loop and B2B growth loop examples.
How do these archetypes compare to AARRR and the HEART framework?
Growth loops are operating systems. AARRR (Pirate Metrics) and HEART are measurement frameworks. They answer different questions.
- AARRR (Acquisition, Activation, Retention, Referral, Revenue): a funnel-stage measurement model. Useful for cohort tracking, not for compounding strategy.
- HEART (Happiness, Engagement, Adoption, Retention, Task success): Google's UX measurement framework, focused on product quality.
- Growth loops: closed-system models that describe how growth compounds, not what you measure stage-by-stage.
A real growth team uses all three: pick a loop archetype, measure it with AARRR-style cohort metrics, and validate product quality with HEART.
For the deeper comparison, see AARRR vs HEART vs growth loops.
| Archetype | How it compounds | Best examples | Failure mode | Leading indicator |
|---|---|---|---|---|
| 1. Viral / Invite | Each user invites N more users inside the product | Figma, Dropbox | K-factor stuck below 0.3 | Invites sent per active user (weekly) |
| 2. Content / SEO | Pages rank, drive sign-ups, who produce more pages | Pinterest, Tripadvisor | Thin pages get deindexed by Google updates | Indexed pages with Top-10 rankings |
| 3. Paid / Reinvestment | LTV from paid users funds more paid acquisition | HelloFresh, Wish (early) | CAC payback exceeds 12 months | LTV:CAC ratio and payback period |
| 4. Sales / Expansion | Closed-won revenue funds more reps, who expand inside accounts | Salesforce, HubSpot | NRR drops below 100% during expansion push | Net Revenue Retention (NRR) |
| 5. UGC / Marketplace | Each side of the market attracts the other via user-generated supply | Airbnb, Pinterest | Liquidity gap on supply or demand side | Supply/demand match rate per geo |
| 6. Data-Network | Usage generates data that improves the product, attracting more usage | Google, Waze | Cold-start without enough seed data | Model accuracy lift per 10x data |
| 7. Ecosystem / Integration | Partners build on the platform, raising switching costs and retention | Salesforce AppExchange, Shopify | Partner economics break, ecosystem stalls | Active integrations per customer |