Building a community platform for the Indonesian market

I’m building Ayoayo.id with a partner, with one simple idea in mind: give Indonesian creators a place to run paid communities, sell courses and other digital products, and handle local payments without having to stitch together multiple tools.

We often describe it internally as “Skool for Indonesia.” It’s an easy way to explain the concept, but it doesn’t quite capture what we’re actually building.

And that gap between the shorthand and the real product is where this case study begins.

The situation

Indonesian creators are already selling. That’s the first thing to understand. This isn’t a market waiting to be taught how to monetize. Indonesia has close to 12 million content creators, the most in Southeast Asia, and its digital economy reached US$90 billion in GMV in 2024, also the largest in the region.

The pattern we kept seeing goes like this: a creator builds a free lead magnet, collects WhatsApp contacts, sells a one-time class, then funnels buyers into a free WhatsApp or Telegram group where the next product gets sold. Every step of that works. What doesn’t work is the middle: matching who paid, opening access, reminding people whose access expired, and knowing how much actually came in this month. That part lives in spreadsheets and bank-transfer screenshots, checked one by one.

I should be upfront about the evidence here. This pattern comes from observation and our own network, not structured interviews. Those haven’t happened yet, and I’ll come back to that at the end.

The obvious question

If the problem is “creators need a community platform with payments,” that product already exists. Skool, Circle, Patreon, Whop, Teachable. Skool alone hosts over 170,000 communities. These are mature products built by teams much bigger than ours. Why build a new one?

Our honest early answer was “because ours will be in Indonesian and cheaper.” Which is not a reason to build a company. The real answer only surfaced when my co-founder mapped the payment architecture, and it turned out to be structural, not cosmetic.

The wall

Here’s the finding that shaped everything: QRIS and Indonesian e-wallets are push payments. The buyer has to approve every single transaction, every time. There is no recurring mandate: no mechanism to charge someone automatically next month. Only cards can do that.

And cards barely exist here. Credit card penetration in Indonesia has sat at around 5%, against 35% in Thailand and 30% in Malaysia, roughly 14.4 million cards in a country of 280 million people.

Now look at what Skool, Patreon, and Circle are actually built on. Not just card payments: automatic recurring billing. That isn’t a feature; it’s the foundation of the business model. Predictable revenue, low-friction retention, members who never have to re-decide. Remove auto-charge and the whole model changes shape.

Recurring e-wallet billing does technically exist. Xendit positions itself as the first gateway in Southeast Asia to offer it. But “first to offer” tells you it’s something that has to be purpose-built on top of the rails, not how the rails behave by default. And with average e-wallet top-ups around Rp140,000 a week (an IPSOS figure from 2020, so read it as directional), balances are thin enough that auto-debits would fail constantly anyway.

For the majority of Indonesian buyers, “subscription” can only ever mean: pay for a period, get a reminder, decide again.

Which means you can’t translate Skool into Bahasa Indonesia. You have to redesign how the money moves.

A market with hard edges

The rails don’t just lack auto-charge. They have limits, and the limits are regulation, not preference.

QRIS is capped at Rp10 million per transaction. Registered e-wallets max out at Rp20 million in balance and Rp40 million in monthly transactions. So a Rp45 million bootcamp, or an Rp11 million year of tutoring, literally cannot go through the cheap channels. High-ticket products need bank transfer. That’s not a preference we can design around. It’s a hard edge the product has to accommodate.

Fees are set by the regulator too. QRIS costs regular merchants 0.7%; micro-merchants pay nothing on transactions under Rp500,000 (and 0.3% above that). Which produces an uncomfortable truth: a creator running their own micro-merchant QRIS pays less on small transactions than we do as a platform. We can’t win on transaction cost. Whatever we charge has to be justified by what sits around the payment (access automation, reminders, bookkeeping), not by cheaper rails.

One more door closes here: Bank Indonesia prohibits surcharging QRIS fees to buyers. So “just pass the fee to the customer” is off the table on exactly the channel where it would matter most.

The tools that exist

Three categories of tools serve this market today, and each one misses in its own way.

Course marketplaces take roughly half. Kelas.com’s own mentor page offers creators “up to 50%” of each sale, meaning the platform keeps up to the other half. That’s the price of traffic and brand. For a creator who already has an audience, it’s paying heavily for something they don’t need.

Local monetization tools are far cheaper. But look at what each one is actually built around:

Fees verified from each platform’s own pages, August 2026. They change often. The date matters.

Read the third column and the pattern is obvious: these are payment tools first. Communities and courses, where they exist at all, are attachments. None of them is built around the thing our target creator actually runs: an ongoing community with courses inside it and members whose access expires.

International platforms are the mirror image: the right product on the wrong rails. Skool, Circle, Patreon, Whop, Teachable. Not one of them accepts QRIS or Indonesian e-wallets. Everything runs through Stripe and cards, and Gumroad can’t even connect Stripe payouts to Indonesia at all.

The gap is specific: no Indonesian platform combines community, courses, and local payments in one integrated product. That gap exists because of the payment wall, not despite it.

The US$23 million warning

There’s a version of this story where someone well-funded already tried. TipTip raised US$23 million (the largest war chest of any Indonesian creator monetization startup), with a US$10 million seed in 2022 and a US$13 million Series A led by East Ventures the same year. By 2026 they had pivoted out of creator monetization entirely, into entertainment ticketing, and only reached profitability after the pivot.

The CEO of KaryaKarsa has been open about why the Patreon model struggles here: card and PayPal penetration, and the subscription habit itself. Indonesian consumers are still getting used to subscribing to anything at all. KaryaKarsa was deliberately built on e-wallets and one-off purchases instead.

None of this is proof. But it’s the strongest signal available, and it points the same direction as the technical finding: recurring billing is not this market’s native behavior.

The language the market already speaks

Here’s the part that made the product click. Indonesians already pay upfront for fixed periods. Constantly. It’s just not called a subscription.

Gyms sell 3, 6, and 12-month packages where the per-month price visibly drops with duration. Tutoring chains like Brain Academy sell by school year, at Rp11 million and up. Kampung Inggris language camps sell two-week to six-month packages in fixed cohorts. Coding bootcamps run Rp45 million cohorts with set start dates. Nobody had to be taught this format. It’s the native pricing language of Indonesian education and services.

So the shape was sitting there the whole time: not monthly auto-renewing subscriptions, but duration packages (pay for a period, get reminded before it ends, decide again). The reminder does the work the recurring mandate can’t.

What we built

The product thesis, in one sentence: not “Skool but Indonesian”, a paid-community platform designed from scratch for a market whose payments can’t auto-charge.

Members pay for a period. The system tracks every member’s access window and sends reminders before it expires. Where someone pays by card, auto-renewal can quietly work as expected. Where they pay by QRIS or e-wallet (most of the market), the reminder flow carries the renewal. We removed the “prepaid vs recurring” choice from the creator’s interface entirely, because the billing mechanic isn’t a business decision the creator makes; it’s a consequence of how each member chooses to pay.

The positioning follows from the same logic. Ayoayo is not a payment gateway. Payment is the trust layer, not the headline. It’s not a replacement for WhatsApp or Facebook groups. It’s the management layer on top of communities that already exist, because asking creators to migrate their audience is the most expensive ask in the world. And it’s not a marketplace. Creators bring their own audience, so it makes no sense to pay half your revenue for traffic you don’t use.

Why this is hard to copy

Every early-stage story owes an answer to one question: what happens when someone bigger notices? Ours is unusually concrete, because the barrier isn’t a feature. It’s the payment layer this whole case study has been about.

For the international platforms, supporting Indonesia isn’t a localization ticket. It means integrating QRIS and local e-wallets, rebuilding a billing engine around payments that can’t auto-charge, and unwinding a decade of product decisions that assume silent renewal. Skool’s pricing, retention mechanics, and checkout all lean on the recurring mandate. Remove it and you’re not translating the product. You’re redesigning it. For a company serving 170,000+ communities that work fine on cards, that’s a lot of work for one market. Payment rails are a boring moat. Boring moats hold.

For the local tools, the gap runs the other way. They have the rails but not the product. Becoming a community-and-courses platform means rebuilding their core, not adding a feature. Some will try. Mayar is closest in ambition. But being built around this use case from day one is a real head start, not a slogan.

And against the marketplaces, the math argues for itself: a creator who brings their own audience and keeps roughly 90% doesn’t need to hand over up to half for traffic they never use.

I want to be careful not to oversell this. Integration work slows competitors down; it doesn’t stop them. What the payment wall buys us is time, and the durable advantage has to be built in that time, out of the things that compound: creators who run their business here, members who renew here, and a product that understands duration packages, cohorts, and reminder-driven renewal better than anyone retrofitting them.

What we haven’t validated

A case study that only contains evidence agreeing with itself is weaker, not stronger. So, plainly:

• No structured creator interviews yet. The pain points above come from observation, not confirmed research. This is our biggest gap, and it’s next.
• No published churn or retention data exists for Indonesian paid communities. We checked. Every available benchmark is Western, and borrowing them would be dressing up a guess.
• We don’t know whether “decide again each period” hurts retention compared to silent auto-renewal. That’s an interview question, not a spreadsheet question.

My honest read is that the structural argument is solid (the payment wall is real, regulated, and verified), but the product bets stacked on top of it are still bets. That’s pretty much the whole job right now: turning the second half of this case study from reasoning into evidence.

All market figures were verified against primary sources (Bank Indonesia regulations, platform pricing pages, e-Conomy SEA 2024, GroupM-GOAT) in August 2026. Competitor fees change often; treat them as a snapshot with that date attached.