Every famous PLG company you can name grew on the back of one or two well-engineered loops — not a hundred growth hacks. The good news: growth loops come in recognizable patterns you can adapt. Here are the seven that work best in B2B SaaS, each with the real companies that made the pattern famous, the mechanics, and the conditions under which it works for you.

1. The collaboration loop (Slack, Figma)

Mechanic: the product is more valuable with colleagues in it, so users invite colleagues as a natural part of using it. New users experience the same pull and invite more.

The Slack version: a message to a colleague who is not yet on Slack is an invitation by construction. Every channel created is a reason to add more of the org. The product's core action is the viral action — that is the gold standard.

Works for you if: your product has genuine multi-player value (shared context, handoffs, review flows). It fails when collaboration is bolted on — users can smell an invite prompt that serves you rather than them.

Key metric: invites sent per new activated user, and the invitee activation rate.

2. The artifact loop (Calendly, Loom)

Mechanic: using the product produces an artifact that non-users receive — a booking link, a video, a shared report. The artifact works for the recipient and carries the product's signature.

The Calendly version: every scheduling link sent is a live product demo delivered at the exact moment the recipient has the same problem (scheduling). Illustratively, if each user sends links to a handful of new people per month and even a small percentage sign up, the loop pays for itself forever.

Works for you if: your product's output travels — via email, chat, or link — to people outside the account. Design the recipient's landing experience as carefully as your homepage.

Key metric: artifact views by non-users, and view-to-signup conversion.

3. The storage/sharing referral loop (Dropbox)

Mechanic: users get more of the product's core resource by referring others; the referred get a bonus too.

The Dropbox version: the famous two-sided space bonus — invite a friend, both of you get extra storage. It worked because the reward was denominated in product value (storage), not cash, so it attracted people who wanted the product. Dropbox has publicly credited referrals with dramatically accelerating its early signups.

Works for you if: you have a metered resource users actually want more of (storage, credits, seats, runs). Cash rewards attract mercenaries; product rewards attract users.

Key metric: referral participation rate, and 90-day retention of referred users versus baseline.

4. The content/UGC loop (Notion, Canva)

Mechanic: users create content with the product; some of it is public and indexable; search traffic on that content converts into new users who create more content.

The Notion version: thousands of public Notion pages, templates and guides rank for long-tail queries. Every public page is a landing page Notion never had to write. The loop is slow to ignite and nearly impossible for competitors to copy once running.

Works for you if: usage produces content with standalone value that users are willing to publish. You must invest in making publishing easy, fast to index and attractive to share.

Key metric: indexed public pages, organic sessions on user-generated pages, and their signup conversion.

5. The template/marketplace loop (Notion again, Webflow)

Mechanic: power users package their work as templates; templates attract new users searching for solutions; some new users become template creators.

The Notion version: the template gallery turned customers into a distribution force — a creator publishing a project tracker template does Notion's marketing with more credibility than Notion could buy. Webflow's showcase and template marketplace do the same for sites.

Works for you if: your product has a configuration layer worth sharing (workflows, boards, designs, dashboards). Give creators attribution and, ideally, an incentive — status or revenue share.

Key metric: template installs, and install-to-active-account conversion.

6. The integration loop (Zapier-style ecosystems, Slack apps)

Mechanic: your product embeds into the tools your users already live in; each integration surface exposes you to that tool's audience, and integration directories send qualified traffic both ways.

The Slack app directory version: thousands of tools built Slack integrations to reach Slack's users — and each integration made Slack itself stickier and more discoverable. For a smaller SaaS, being listed in the marketplaces of bigger platforms is distribution you do not pay for per click.

Works for you if: your data or actions are more valuable connected. Prioritize the two or three platforms where your ICP already works.

Key metric: signups attributed to integration directories, and retention lift of integrated accounts (typically substantial — integrated users have more to lose by leaving).

7. The paid/revenue loop

Mechanic: revenue from cohort N funds acquisition of cohort N+1 at a known CAC and payback. Not glamorous, entirely compounding — if payback is fast enough.

Mechanics that make it work: CAC payback under 12 months (under 6 is excellent, illustratively), stable channel performance, and net revenue retention above 100% so cohorts grow while they repay.

Works for you if: you have at least one channel with reliable unit economics. The loop breaks silently when channel costs creep up — measure payback monthly, per channel.

Key metric: CAC payback period and the ratio of new MRR to acquisition spend.

How to evaluate a loop before you build it

For any candidate from the seven patterns, pressure-test it on paper first:

Estimate each edge from data you already have. What % of current users naturally do the seed behavior (share, publish, integrate)? How many non-users does each act reach? What is a defensible conversion rate for that exposure (use your current landing-page conversion as a ceiling, not a floor)?
Multiply the edges into a coefficient and note the cycle time. A coefficient under ~0.05 with a monthly cycle will be invisible in your growth curve; either find a stronger edge or a different pattern.
Check the retention precondition. Every pattern above assumes the arriving users stick. Run the projected loop against your actual cohort curves — a loop feeding a 20% month-3 retention product mostly recruits future churn.
Estimate the build honestly. Content and template loops need quarters before compounding shows; collaboration and artifact loops can show signal in weeks. Match the loop's time-to-signal to your runway and patience.

Only after the paper model survives these four checks does the loop deserve engineering time.

Choosing and stacking loops

Three rules:

Match the loop to your product's natural exhaust. Collaboration products get collaboration loops; output products get artifact loops; configuration-rich products get template loops. Forcing a mismatched loop produces spam.
One loop at a time. Instrument it edge by edge (participation rate × reach per participant × conversion), get the coefficient above roughly 0.2, then add the next.
Stack loops with different cycle times. Slack stacked a fast collaboration loop on top of a slower integration ecosystem; Notion stacked fast sharing on slow content. Fast loops give momentum; slow loops give durability.

Before committing engineering time, model the candidate loop: estimate each edge from your current data, simulate the compounding over 12 months, and check whether the realistic (not optimistic) scenario moves your growth curve. That modeling step — drawing the loop, attaching probabilities, running the simulation — is exactly what Growth Pilot's visual loop builder and Monte-Carlo simulator were built for, so you pick your Dropbox moment on evidence rather than envy.