# Growth Loops for SaaS > What a growth loop is, the four loop types that work in SaaS, how to compute loop output and cycle time, and why funnels stop compounding after Series A. Source: https://saas-marketing.net/guides/growth-loops/ Topic: SaaS Growth Marketing Type: guide Published: 2026-09-11 Last updated: 2026-09-11 Publisher: SaaS Marketing (saas-marketing.net) License: CC BY 4.0. Quote or republish with attribution and a link to https://saas-marketing.net/guides/growth-loops/ ## Short answer A growth loop is a closed system where the output of one cycle becomes the input of the next, so growth compounds instead of requiring fresh spend each month. It has four parts: an input such as a new user, an action that user takes, an output that action produces, and a reinvestment step that feeds the output back as a new input. Funnels are linear and end at conversion. Loops restart. ## Key takeaways - A loop compounds only if the output of one cycle is mechanically reinvested as the input of the next cycle. - Loop output and cycle time are the two numbers worth optimising: a 0.4 coefficient at seven days beats 0.8 at ninety. - Most B2B SaaS products only ever get a weak content loop, and that is an acceptable outcome if you run it well. - Figma, Miro, Loom and Calendly all grew on collaboration loops where sharing is the core action, not a bolt-on feature. - A referral programme attached to a product with no natural sharing moment is a discount, not a loop, and it will not compound. --- The funnel isn't wrong, it's just finished. It describes one person's path from not knowing you to paying you, and then it stops. Nothing in the model explains why next month should be cheaper than this month. A loop does. The difference is a single structural property: the output of one cycle becomes the input of the next. ## Loop anatomy: four parts, one of which everyone forgets Every working loop has an input, an action, an output and a reinvestment step. The first three are easy to identify. The fourth is where most "loops" turn out to be funnels with ambition. Take Loom. Input: a new user signs up. Action: they record a video and share the link with colleagues. Output: viewers land on a Loom-hosted page. Reinvestment: some fraction of those viewers sign up to record their own video, which becomes a new input. That last arrow is mechanical, built into the product, and requires no marketing spend per cycle. Now take a typical B2B product with a "refer a friend, get $50" banner in the settings page. Input, yes. Action, sort of. Output, occasionally. Reinvestment? Only if someone remembers the banner exists. That's not a loop, that's a promotion with a low redemption rate. Ask: if we stopped all marketing spend for ninety days, which of these mechanisms keeps producing new users? Whatever survives is a loop. Everything else is a channel, and channels need feeding. ## The four loop families that work in SaaS ### Collaboration loops The strongest family, and the hardest to retrofit. The loop works because using the product normally exposes non-users to it. Figma's multiplayer cursor, Miro's shared board link, Calendly's booking page carrying a small link at the bottom of a page thousands of people see each week. What makes these work is that the sharing moment is the core action, not an added one. Nobody at Figma has to persuade a designer to invite a stakeholder. Getting feedback requires it. The math: loop output equals invites sent per user times acceptance rate times activation rate of the accepted. A B2B collaboration product sitting at 0.3 to 0.7 is doing well. Above 1.0 and you have a consumer product that happens to bill businesses. The [Viral Coefficient Calculator](/calculators/viral-coefficient/) runs the arithmetic with your own numbers. ### Content loops The family most B2B companies can actually build. User or product activity generates pages, pages attract search traffic, traffic becomes users, users generate more pages. Zapier is the textbook version. Every app added to the platform spawns integration pages for every pairing with existing apps, and that surface area grows combinatorially. Canva does the same with its template index. G2 runs it on reviews. The trap: most companies build a blog and call it a content loop. A blog is a content programme. Editorial output doesn't scale with user count, so there's no reinvestment arrow. The test is whether adding a thousand users mechanically produces new indexable surface without a writer touching anything. ### Paid loops Revenue from acquired customers funds the next round of acquisition. Structurally a loop, but the cycle time equals your CAC payback period, which for most B2B SaaS is long. Benchmarkit's work puts median CAC payback around 16 months. A loop with a sixteen month cycle compounds so slowly it barely feels like a loop at all. Paid loops work when payback drops under about twelve months and gross margin is healthy. Below that, you're not compounding, you're financing growth with the balance sheet, and that's a different conversation with a different risk profile. ### Sales-driven land and expand Land a small team, expand to adjacent teams, use the resulting revenue to fund more sales capacity. This is how most enterprise SaaS actually compounds, and it depends entirely on net revenue retention. Above roughly 115 percent NRR the expansion motion pays for meaningful new capacity. Below 105 percent it doesn't. The marketing contribution here is customer-facing content that helps champions sell internally, which is why this loop lives closer to customer marketing than to demand gen. ## The two numbers worth optimising Loop output and cycle time. Almost everything else is a distraction. Loop output is how many new inputs one input generates across a full cycle. Cycle time is how long that takes. Compounding rate is roughly output raised to the power of cycles per period, which means cycle time has an exponential effect while output has a linear one. A loop with output 0.4 and a 7 day cycle runs about 52 times a year. A loop with output 0.8 and a 90 day cycle runs about 4 times. The first one produces dramatically more compounding despite half the per-cycle output. Shortening the cycle is also usually a product or onboarding fix, which is cheaper than raising the coefficient. Practical levers on cycle time: shorten time to the sharing moment in onboarding, remove sign-up friction for invited users, make the shared artifact viewable without an account. Figma's decision to let anyone view a file link without signing up cut a large chunk of cycle time out of their loop. Faster time to value shortens every loop that depends on activation, which is why [Time to Value (TTV)](/glossary/time-to-value/) sits upstream of loop performance. ## Why most B2B products only get a weak content loop Here's the uncomfortable part. If your product is used by one person, produces no shareable artifact, and serves a market of four thousand companies, you will not build a viral loop. No amount of referral programme design changes that. What you can build: a content loop seeded by product data or customer activity, and a strong land-and-expand motion. That's it, and it's fine. Plenty of $50M ARR companies run exactly this combination. The mistake is spending two quarters of engineering time trying to manufacture virality. Bolting a referral programme onto a product with no natural sharing moment produces a discount with extra steps. It costs margin, generates a trickle of low-intent signups, and the team concludes loops don't work in B2B, when what actually happened is they picked the wrong family. What to do instead if you're in this position: - Audit whether any user-generated data could become public pages with consent - Build integration or comparison pages that grow with your partner ecosystem rather than with your editorial calendar - Instrument the expansion path properly and treat seat growth as the loop - Accept that paid is a channel, fund it as a channel, and stop calling it a loop in board decks [12 SaaS Growth Loop Examples](/guides/growth-loop-examples/) has fuller teardowns if you want to see which family your product could plausibly fit. ## Instrumenting a loop in product analytics You cannot optimise a loop you cannot see, and the default analytics setup will not show you one. Loops are cross-user relationships, and most event tracking is per-user. **Instrumenting a loop end to end** That last step matters more than it looks. The single most common way a legitimate loop gets abandoned is a premature measurement window. Write the review date into the experiment brief before you start. [Growth Experiment Brief Template](/templates/experiment-brief/) has a field for exactly this. ## When to stop and just run a funnel Honest tradeoff: loop work is slower to show results than channel work and harder to explain to a board. If you're at $800K ARR with eleven months of runway, building a content loop that pays off in month fourteen is the wrong call. Run outbound, run paid, close deals, live to fight. Loop investment makes sense when you have eighteen months of runway or better, an existing base large enough to seed the loop, and a product where a plausible reinvestment arrow already exists in some weak form. Look for a loop you can strengthen rather than one you must invent. The [B2B SaaS Growth Strategy by Stage](/playbooks/b2b-saas-growth-strategy-by-stage/) playbook covers which motion fits which ARR band. ## What to do this week Draw your current loop on one page: input, action, output, reinvestment. If the reinvestment arrow is dotted or hypothetical, you've found the work. Then pick the single shortest-cycle candidate and instrument it properly before changing anything. The [SaaS Growth Model Template](/templates/growth-model-spreadsheet/) will let you model what a given output and cycle time compounds to over eighteen months, which is usually the moment a team realises cycle time is the lever. For definitions and the broader context, [Growth Loop](/glossary/growth-loop/) covers the term precisely, [Product Led Growth Examples](/examples/product-led-growth-examples/) shows loops running in real products, and the [SaaS Growth Marketing](/saas-growth/) hub connects loops to the rest of the growth function. ## Frequently asked questions ### What is a growth loop? A growth loop is a self-reinforcing system with four stages: an input like a new user or a dollar of revenue, an action that input takes, an output that action produces, and a reinvestment mechanism that turns the output back into a new input. The defining property is that running the loop once makes the next run cheaper or larger, which is what produces compounding. ### What is the difference between a growth loop and a funnel? A funnel is linear and terminates. You pour in awareness, some fraction converts, and the process ends at purchase. To grow next month you must pour in more at the top. A loop routes the output back to the input, so each cycle seeds the next. Funnels are useful for describing a single conversion path. Loops describe why a business compounds. ### What is loop cycle time and why does it matter? Cycle time is how long one full pass of the loop takes, from a new input entering to that input generating a new input. A loop with a coefficient of 0.4 and a seven day cycle compounds far faster than one with 0.8 at ninety days. Shortening cycle time is usually easier than raising the coefficient, and it is where most teams should start. ### Do growth loops work for B2B SaaS? Yes, though rarely the viral kind. B2B products that involve collaboration get genuine invite loops: Figma, Miro, Loom, Calendly. Everything else usually gets a content loop or a sales-driven land and expand loop. A vertical product selling to twelve hundred possible buyers will never compound virally, and should optimise expansion and referral quality instead. ### What is a content loop? A content loop uses user activity or product data to generate pages that attract new users, who then generate more pages. Zapier's integration pages, Canva's template index and G2's review pages all work this way. The reinvestment step is what makes it a loop rather than a content programme: new users must mechanically create the next batch of indexable surface. ### How do you measure a growth loop? Instrument four events: loop entry, the core action, the output creation, and the new input attributable to that output. Then compute loop output per input and the median time between entry and new input. Product analytics tools like Amplitude, Mixpanel or PostHog handle this if you define the events before you build the report, not after.