UGC loop: Your users build your SEO moat

Notion passed 1 million users in 2020, 20 million by late 2021, and 100 million in 2024.¹ Running alongside that is a template gallery its own customers filled: thousands of published, duplicable setups for jobs Notion never wrote a page about.

Notion has never said how much of that growth its users’ templates drove, and the “95% organic traffic” figure everyone quotes does not survive checking.² The mechanism is visible without it.

A shared link is used up the moment someone clicks it. A template published years ago is still ranking, still being duplicated, and still recruiting users who have never heard of you.

Why the UGC loop compounds: the artifact outlives the interaction

Most viral loops are transactions. Someone sends a scheduling link, a stranger books a slot, the loop completes and the link is worthless. The loop only spins as fast as your users keep using it.

A user-made artifact behaves differently, for three reasons.

It is searchable. “Product roadmap template” is a query with steady demand, and the page answering it was written by a customer, for free, three years ago. Every specific use case your users have is a long-tail query someone else is typing.

It is proof. A stranger evaluating your product usually has to imagine what they could build. A template collapses that: here is a working thing, made by someone with your job, that you can have in one click. Nothing in your marketing does that as well as a customer’s actual work does.

It gets better without you paying for it. The hundredth template is more useful than the tenth, because coverage of the long tail improves. Programmatic pages compound too, but you carry their production cost. Here someone else does.

The cost of that leverage is control. You do not decide what gets made, how good it is, or what it says about your product.

What is a UGC loop, and how is it different from content marketing?

A UGC loop (user-generated content loop) is a PLG acquisition pattern where the things your users make become the content that attracts new users. Templates, designs, databases, public docs. You are not producing marketing; your customers produce it as a by-product of getting their own work done, and it accumulates.

The distinction that matters is who makes the content and how long it lasts:

Content marketing has a treadmill problem: stop writing and the channel decays. A UGC loop has the opposite property. Every user who solves their own problem in public leaves an asset behind, and the library gets larger whether or not anyone at your company does anything that quarter.

Does your product produce anything worth publishing?

A UGC loop is a property of your output, not a growth tactic you can adopt. Which is why the two obvious moves both fail: adding share buttons to a product whose output nobody wants to show, and paying people to create. Paid creation produces volume at exactly the quality that earns the payment, which is the thin-content problem below, bought rather than accumulated.

The products where this already runs are easy to name, and worth checking yourself against: design tools (Canva, Figma), site builders (Webflow), workspace and database tools, documentation tools, and form builders. If your product sits next to one of those on a comparison page, the loop is available to you whether or not you are using it.

Two ways the artifacts get found

Notion: a gallery, so the work is findable

Notion’s product is deliberately unopinionated, which means users build wildly different things in it, which means the raw material for a library exists whether or not anyone curates it. The template gallery turns that raw material into a discovery surface: publishable, browsable, and one click to duplicate.

There is a second, quieter loop in products like Airtable, the spreadsheet-database hybrid whose pricing troubles appear earlier in this handbook and whose acquisition loop was never the problem: participating in someone else’s shared artifact requires an account of your own. That turns every invitation to work together into a signup, without anyone publishing anything.

The mechanism worth copying in the gallery version is the duplication step. Finding a template is worth little; using it requires an account, and the gap between “I want this” and “I have this” is where a UGC loop is won or lost. Every extra step between seeing the artifact and using it costs you conversions, and a signup wall costs the most.

Canva: the output leaves the building on its own

Canva runs both loops at once. It has one of the largest template libraries in software,³ and its artifacts also travel independently of that library: a social post, a deck, a flyer, made specifically to be seen by people who are not Canva users. The question they provoke is “how did you make that?”

The second loop is the one worth studying, because it needs no curation to work. What makes it run is something upstream of both: Canva lowered the skill floor until people who cannot design produced output worth showing anyway. A loop that depends on your users making something impressive only works if ordinary users can. Widening the set of people capable of producing a shareable artifact does more for the loop than any gallery feature, because it widens the supply at the source.

The transferable point is that a gallery matters most when your output stays private by default. If what your users make is already going somewhere public, curation is a bonus rather than the mechanism, and the loop is stronger either way, because the artifact arrives with a person’s endorsement attached rather than sitting in a directory.

Where UGC loops go wrong: the library turns into a liability

Loops rarely stop. They keep running while the output quietly stops being worth anything, which is harder to notice and more expensive.

Every user-made library eventually accumulates duplicates, abandoned drafts, near-identical variations of the same idea, and material made to game a marketplace’s ranking rather than to solve a problem. At small scale nobody notices. At large scale it becomes thin content on your domain, at volume, which is close to the profile Google’s scaled content abuse policy exists to catch.⁴ The policy turns on whether pages exist primarily to manipulate rankings, so one abandoned duplicate is not a violation. The risk is blunter than that: thin material at volume, sitting on your domain.

There is a second, quieter cost. Whatever your users publish becomes what strangers think your product is for. If the top twenty templates for your tool are all one narrow use case, that use case is now your positioning, whether or not you chose it.

Moderation is therefore a growth function, not a hygiene one. Deduplicate aggressively, remove what does not work, and be willing to shrink the library. A hundred templates that each solve a real problem outrank a thousand that mostly do not.

Is your loop actually working?

Your own gallery answers the question in three queries.

1. Do users arriving from a user-made artifact activate better than your other traffic? They should, considerably. They arrived with a working example of the product doing their job. If they don’t, the artifact is attracting the wrong people, or the duplication step is losing them.
2. What share of your library gets used at all in a quarter? Sort by duplications. If most of the library is never used, you have a thin-content problem accumulating rather than a moat.
3. Is the library growing faster than your user base? If it isn’t, the loop is not compounding; you just have customers who occasionally publish. The whole premise is that creation outpaces you.

Read the gallery you did not know you had

1. Look at what your users already publish, outside any surface you built. Search your product’s name plus “template”. If people are already sharing their work in public, the loop exists and you are simply not helping it.
2. Time the path from finding an artifact to using it. Count the clicks and the walls. This single number decides whether a library converts, and it is rarely measured.
3. Give the long tail a home, but only if it needs one. Build the gallery when your output stays private by default. If the artifacts already travel on their own, spend the effort on making them carry attribution instead.
4. Remove the weak templates before you promote the library. Work out what share of your library has ever been used. Removing the dead weight protects the domain and makes the rest easier to find.
5. Feed the people who make the good stuff. Interview them, credit them, put them in front of the audience their work earned. The supply side of a UGC loop is a small number of people doing a disproportionate amount of the work, and they are usually thanked by nobody.
6. Read the top twenty artifacts as a positioning document, because that is what strangers are reading them as.

A template gallery is talk you can see. Much of the most valuable talk about a product happens where you have no account, no analytics and no way to intervene at all. Community-led growth is the discipline of being talked about well in rooms you do not own.

Updated August 2026