Research date: August 1, 2026. Six parallel research threads, roughly 220 verified sources. Every price below was read off the vendor's own pricing page on August 1, 2026, not recalled.

---

Executive summary

There are already more than 30 platforms cheaper than Circle, and "cheaper" is not why anyone switches. The sub-$100 band is packed: Skool at $9, Nas.io at $20.75, SchoolMaker at $29, MemberSpace at $39, Heartbeat at $40, Podia at $42, Uscreen at $49, EzyCourse at $59, Mighty Networks at $79. Skool alone has roughly 170,000 communities. Undercutting Circle on price means walking into a knife fight that Skool already won two years ago, funded by a16z.

But you found something real, and it's one band over. Circle deleted its $49 Basic plan in early 2025, so its floor is now $89/month. And in the entire commercial field, there is no platform under about $199/month that gives a creator a genuine white-label app in their own App Store listing. Below $150, that band is empty. Verifiably, completely empty.

That band is empty for a hard reason, and the reason is the whole decision. Apple's guideline 4.2.6 and Google's white-label policy together make each branded app a per-customer human operations job: a separate developer account, separate legal and banking verification, separate review cycles, separate annual SDK migrations, and on Google, a mandatory 12-testers-for-14-days closed test for every new personal account. At 40 customers that's 40 to 120 days a year of unpaid, unskippable compliance work for one person. That is why Circle, Mighty Networks, and Disciple all sell branded apps as a high-touch enterprise upsell at $399 to $2,500 a month, and why nobody sells them cheap.

The honest verdict: "cheaper Circle with branded apps for everyone" has a 5 to 8% chance of becoming a real business, because the two halves of the pitch fight each other. A focused vertical community platform with one compliant shared app and web-only payments has a 25 to 35% chance, which is a genuinely respectable number for a solo founder. The plan below is built for the second shape.

---

What's actually true about Circle

Pricing, verified August 1, 2026

Add-ons that inflate the real price on Professional: custom profile fields $49/mo, branded sender email $40/mo, extra admins $10/mo each, extra spaces $20/mo per 10, extra live attendees $20/mo per 100, extra live hours $50/mo per 10. A Professional customer who wants profile fields, branded email, and two extra admins pays $198/month, which is Business money without Business features.

The tier history matters

Wayback snapshots of circle.so/pricing tell a clean story:

Circle stranded its own low end and moved upmarket. Professional and Business haven't changed price in over two years. What changed is that the cheap door closed and the expensive door became a sales call. That's a deliberate move toward taking a cut of creator revenue rather than selling seats, and Circle Studios (a revenue-share partnership program) confirms the direction.

Circle just changed what it is

On June 16, 2026, Circle announced "Eclipse" and rebranded as the first AI-native community platform. It reached all customers on July 29, 2026. That includes Circle AI (a conversational agent with memory and 50+ named skills that will scaffold spaces, configure paywalls, build courses, and draft launch plans from a sentence), Circle Discover (a community marketplace), Circle Studios, a rebuilt visual course builder, and a unified Inbox. AI Agents and AI Workflows are Plus-only, with AI Actions metered at 1 credit per comment and 5,000 credits a month included.

This is six weeks old. Anyone building a feature-clone of Circle right now would be cloning Circle-of-2025 and underpricing a product Circle is no longer selling.

Complete Circle feature inventory

Marked with a star are the parts that are genuinely hard to build.

Spaces and structure. Post/discussion spaces, chat spaces, event spaces, course spaces, member directory. Space groups for grouping and access control. Public, private, secret, and paid space access. Personalized member feed and navigation. No space templates or copy-space, so settings get redone per space (a recurring G2 complaint).

Posting. TipTap rich text editor. Video, audio, files, embeds. Threaded comments, reactions, mentions. Scheduled and pinned posts. AI summaries of long threads. SEO-friendly public pages.

Live streaming and rooms. ⭐ Live streams with 100 to 200 attendees plus 6 co-hosts. Live rooms (everyone on camera) with 15 to 20 participants. Recordings auto-saved and shareable. Automated transcriptions on Business and up. Public webinars shipped May 2026. Works on iOS with camera and mic.

Courses. ⭐ Structured, scheduled, and self-paced course types. Sections and lessons, visual learning paths. Drip by section, either relative to each student's enrollment or on fixed calendar dates, with locked lessons. Auto-graded single-choice and multiple-answer quizzes with a minimum passing grade that gates the next lesson. Per-lesson progress tracking. Video-first player. Circle AI drafts whole curricula. No native completion certificates (requires Accredible via Zapier), and progress analytics are shallow: completion yes or no, not time-on-task.

Events. Branded event pages, RSVPs, calendar invites, reminders targeted by space and member tag, recurring events, recordings that become searchable content. RSVP can be restricted to specific access groups while the event stays publicly visible. No native third-party calendar embed in an Events space.

Chat. ⭐ One-to-one DMs, group messages, persistent chat spaces. Real-time across web, desktop app, iOS, and Android. Welcome DMs sent personally or by AI. Automatic peer introductions. Per-chat moderators, message edit and delete, search and archive. Admins cannot read member DMs.

Gamification. Points for likes and contributions, levels with automated unlocks, leaderboards, badges, reward unlocks (exclusive spaces, content, discounts). Available on every plan.

Member management. Rich profiles, searchable directory, member tags (public or admin-only), customizable profile fields, admin and moderator roles, invitations and bulk invites, onboarding welcome messages and pop-ups, member reporting, restrict repeat offenders, approval queues, auto-moderation rules, banned words, activity scores on Business and up.

Monetization. ⭐ Stripe only. Paid memberships, recurring subscriptions, one-time payments, trials, percentage or fixed coupons. Access groups auto-granted and auto-revoked on payment events. Affiliate program support and conversion tracking. Branded checkout. 25+ currencies, all major cards and wallets. Payments analytics: churn, trial conversion, MRR, per-paywall conversion. Free subscription migration service from other platforms. Transaction fee 2% / 1% / 0.5% on top of Stripe's 2.9% + $0.30.

Email and notifications. ⭐ In-app, email, and mobile push. Weekly digest. One-click post-to-email. Branded sender email. Email Hub is a separate product: broadcasts, marketing automations, forms, contact CRM, segmentation, open and click analytics, drag-and-drop builder, sends to non-members. Free under 100 contacts, then priced by contact count, and it's an add-on on both Professional and Business.

Search, discovery, and AI. ⭐ Search with unlimited history. Circle AI on every plan with 50+ skills including Space Scaffolder, Community Restructurer, Course Builder, Paywall Configurator, Gamification Setup, Content Moderator, Business Model Advisor, Workflow Builder, and Analytics Manager. AI Agents on Plus only: up to 10 agents, retrieval over a community knowledge engine, custom tone, conversation memory, and an AI Inbox where a human can pause the agent and take over mid-thread. AI Workflows on Plus only: describe in plain language what content to act on (negative sentiment, spam, topics), then auto-comment, hide, report, DM, or re-tag.

Branding. Custom domain on every plan including Professional. Custom CSS and JavaScript. Color themes and fonts. Removing Circle branding requires Business. Website Builder with 100+ blocks, dynamic blocks previewing events and posts, native lead-capture forms, page analytics, gated content previews. Headless front-end on Plus.

Mobile. The standard Circle Communities app for iOS and Android plus a desktop app, rated 4.87 across roughly 76,000 ratings, with full parity to desktop. Branded white-label app on Circle Plus only.

Analytics. Traffic sources, top-converting pages, growth over time by space/course/paywall, segment behavior filtered by tag and profile field, login patterns, paywall ROI, trial conversion, event attendance and course participation tied to revenue.

Integrations and API. ⭐ Zapier, Integrately, embed widgets, embeddable comments. SSO via Facebook, Twitter, Memberstack, MemberSpace, Teachable, Memberful, WordPress, Outseta, Bubble, Auth0; custom SSO is Plus only. Admin API (REST v1 and v2) on Business and up, 5,000 to 30,000 requests a month with per-request overage. Headless Member API with JWT auth. Data API on Plus only. WebSockets in beta. A hosted Circle MCP server on Business and up, with documented setup for Claude, Cursor, ChatGPT, and VS Code. REST only, no GraphQL.

Automations. Not available on Professional at all. Business gets 20 automations, 20 scheduled workflows, and 20 bulk actions a month. Event-triggered: tag added, paywall event, form completion, RSVP, join. Actions: add or remove from spaces, connect peers, announce, send event reminders, award badges, remove churned members. Visual editor, and Circle AI builds workflows from natural language.

Admin. Bulk import and export, sandbox community on Plus, migration services (payments free on all plans, courses free on annual Business and up, dedicated migration lead on Plus). GDPR, CCPA, PCI SAQ, SOC 2.

Circle's known gaps

1. No free plan and no tier under $89. The clearest wedge in the whole product.
2. No native course completion certificates.
3. Shallow LMS: basic multiple-choice quizzes, no graded written work, completion-only analytics.
4. No automations at all on Professional, only 20 on Business.
5. No API on Professional. Programmatic access starts at $199/mo annual.
6. REST only, WebSockets still beta.
7. 10 live hours a month on Professional is brutally low for weekly calls, and overage is $50 per 10 hours.
8. No space templates or copy-space.
9. 200 GB storage cap on Professional constrains video-heavy communities.
10. Stripe is the only processor.
11. "Missing features" is the single most-cited G2 negative, with 47 mentions.

Company signals

Founded January 2020 by Sid Yadav, Andy Guttormsen, and Rudy Santino. Roughly $33.3M raised across five rounds, including a Tiger Global-led round in December 2021 at a $200M valuation. Sid Yadav publicly stated $20M+ ARR and a $250M valuation in May 2024. Headcount around 344 as of May 2026. 4.5 stars across 501 G2 reviews. Public customers include Miro, Codecademy, Pat Flynn, Ali Abdaal, Johnny Harris, Tiago Forte, and Jay Clouse.

---

The competitive field

Everyone cheaper than Circle

Prices verified August 1, 2026 on each vendor's own page. Annual figures shown where the vendor leads with them.

Also verified: Frond shut down May 31, 2025. Geneva was acquired by Bumble and the standalone app was shut down in September 2025, having generated zero revenue post-acquisition despite a $17.5M price and real user love. Those were the two best-designed independent entrants of recent years, and both are gone.

The price band map

The empty bands, stated precisely:

• Under $30/month with a genuine white-label app: completely empty.
• $30 to $150/month with a genuine white-label app: empty. The floor is roughly $199 (EzyCourse), realistically $399 (Disciple).
• Under $60/month with near-zero transaction fees and a real community feature set: nearly empty, only SchoolMaker and EzyCourse, both small brands.

Open source: there is no Circle clone, and that's a signal

Searched GitHub directly. Every "circle.so alternative" and "skool clone" repo is a 0-to-10-star abandoned side project. After twenty-five years of vigorously maintained open-source community software, with tens of thousands of stars and thousands of contributors, nobody has shipped a credible open multi-tenant creator-community platform with payments.

That's not because nobody thought of it. It's because the hard parts (Stripe Connect with platform fees, App Store operations at scale, tenant isolation, and customer support) are operational businesses, not code, and volunteers don't build operational businesses. Circle's moat is not its feature list. Its moat is that the feature list is the cheap part.

What does exist, verified live on August 1, 2026:

Two genuinely useful surprises here. Discourse bundles working Stripe recurring subscriptions in core, gating group access, in the free self-hosted build. And Discourse is GPL-2.0, which has no network clause, so running it as a hosted SaaS is not "distribution" and would not force you to publish your own source. Almost everything else modern (Forem, Mastodon, Lemmy, Synapse, Open edX, LearnHouse, CourseLit) is AGPL-3.0, which does trigger source disclosure for a hosted service and therefore rules it out for a proprietary SaaS.

Multi-tenancy is the wall. Not one of these was built to host many separate private communities from one deployment with self-service signup. Discourse's own team says multisite is an advanced sysadmin setup they won't support, all sites must share plugins and upgrade in lockstep, and there's a documented CVE involving Stripe API key leakage between tenants in exactly the multisite-plus-subscriptions combination a Circle competitor would need. NodeBB, Flarum, HumHub, and Mastodon are one-community-per-install, full stop.

So the tenant control plane (provisioning, isolation, per-tenant domains, per-tenant branding, per-tenant Stripe accounts, staged upgrades) is 100% net-new work on every option. No fork gives you the actual product.

---

What users actually say

Price is the loudest complaint and it is not why they leave

The most-upvoted "I'm leaving Circle" thread on Reddit was written by a $5,000-a-year enterprise customer, and she later edited it to say she stayed: "I still don't have a better alternative" and "the amount of time and trial and error it took to get to Circle, when I was faced with that prospect again I was just ready to throw in the towel." She had already tried BuddyBoss, Skool, Facebook Groups, and Kajabi.

That's the moat: evaluation fatigue plus migration cost. Not price, not features.

Where people go when they leave, and why

Almost every migration story is Circle to Skool, at $99/month, which is more than Circle's $89 entry plan. The stated reason is always the same:

• "Circle looks wayyyy too complex. Skool is the simplicity I NEED as a mama of 4."
• "I am moving from Mighty Networks and Kajabi! I just need simplicity! I am exhausted from the tech tools."
• "For someone who's an extroverted socializer like myself, too many bells and whistles is a distraction."

People pay more to get less. Cheap is not the axis they move on.

The complaints, ranked by real frequency

1. The gap between advertised and real price. Not "$89 is a lot" but "$89 is a lie." Email Hub extra, admin seats extra, profile fields extra, branded email extra, plus a 2% transaction fee that isn't on the pricing page.
2. Feature-gating that forces upgrades. Bulk-deleting members requires the $649 plan. One customer: "It's basic functionality that you would expect a newly developed app in its beta phase to have."
3. Billing behavior. Multiple independent Trustpilot reports of unannounced increases ($89 to $129 with no notice) and charges after confirmed cancellation ($214.97, twice). Mighty Networks and Teachable are generating identical rage right now.
4. Support quality. "Support is inconsistent at best. Difficult to access." The AI support agent is called "absolutely useless."
5. Complexity and member confusion. The dominant migration driver.
6. Dead communities. Circle communities run around 20% monthly active members; Mighty Networks around 33%. One creator: "I paid 20K to MN for white labeled and then 12K to circle for mobile app. Hardly any engagement."
7. Upsell popups and notification spam inside the paid product.
8. Migration lock-in. Post history, DMs, and activity data don't transfer, and active subscriptions must be re-enrolled by hand.
9. Mobile experience. Repeatedly asked about in reviews, never answered by anyone.

Notice what is not near the top: SEO, white-labeling, and API access. Those are vendor talking points, not user pain.

What people wish existed, and nobody sells

• Full member export including payment history, plan, and tenure. A competing founder named this as the single thing that ties you to a platform, and said it "almost never" exists.
• Pricing that doesn't tax growth. The sharpest strategic argument in the whole corpus: a percentage costs the same at $500/month in sales as at $50,000, while the platform's cost to serve you barely changes.
• Simplicity at Circle's feature depth. The real hole between Skool (too simple, no tiered memberships) and Circle (too complex). People bounce between the two and are unhappy in both directions.
• Notification and email controls that actually work for members. Members complaining to owners about the platform is uniquely corrosive.
• A genuinely mobile-first experience.

Willingness to pay

Small community owners anchor at $50 to $100/month and treat anything above about $130 as an insult unless revenue justifies it. Nobody calls Skool's $99 flat rate unreasonable. They call $89 that becomes $268 unreasonable.

Willingness to pay is not the constraint. Willingness to switch is.

The uncomfortable truth

"Too expensive" is usually code for "I am not making money from this." One verbatim: "I'm about to cancel as well, too expensive and I'm not getting back anything at the moment, more than a year using." A creator earning $3,000 a month never posts that.

A cheaper platform doesn't fix an unprofitable community. It just makes the failure cheaper. If you build a cheaper Circle, you inherit a customer base whose actual failure mode you haven't addressed, and they will churn out of your product for exactly the reason they churned out of Circle: nothing was happening in there.

---

The two things that would kill this

1. The App Store, and it's worse than it looks

Apple guideline 4.2.6, current verbatim wording:

Apps created from a commercialized template or app generation service will be rejected unless they are submitted directly by the provider of the app's content. These services should not submit apps on behalf of their clients and should offer tools that let their clients create customized, innovative apps that provide unique customer experiences. Another acceptable option for template providers is to create a single binary to host all client content in an aggregated or "picker" model.

That leaves exactly two legal architectures:

• Path A, the aggregator app. One binary, your brand, every customer's community inside it. One developer account, one review queue, one SDK migration a year. This is what Circle's and Skool's standard apps do. Fully compliant. The cost: no customer gets their own icon on the home screen, which is the entire reason people pay for branded apps.
• Path B, per-customer branded apps. Each customer opens their own Apple Developer account ($99/yr) and Google Play account ($25), and you're added as a team member. Compliant, and it's what Mighty Pro does. But every branded app becomes a support ticket chain: account creation, legal entity verification, tax and banking forms, certificates and profiles, per-app review cycles, annual renewals.

Google is blunter. Its white-label guidance recommends decentralized accounts because "policy issues with one app can negatively impact all other apps within the same account... this can result in serious account level consequences that can ultimately impact every app within the account (including account suspension or even termination)." Publish 40 near-identical apps under one account and you can lose all 40 at once.

And a concrete new friction point: any personal Google Play developer account created after November 13, 2023 must run a closed test with 12 testers for 14 consecutive days before it can publish. That's a two-week floor plus recruiting 12 real humans on real devices, per customer.

The forced maintenance treadmill. Since April 28, 2026 every App Store Connect upload must be built with Xcode 26 against the iOS 26 SDK, and Apple resets this every April. Google Play requires API 35 by August 31, 2026 and API 36 for updates, and that repeats annually. One aggregator app costs 3 to 6 weeks a year in pure compliance work. Each branded app adds 1 to 3 days at launch plus half a day to a day per annual cycle plus incident time. At 40 branded customers that's 40 to 120 days a year of unpaid, non-differentiating work for one person.

That is the number that ends the "branded apps for everyone" business. It is why GoHighLevel charges $497/month for the same capability and still has a standing customer thread titled "Whitelabel Mobile App is Too Expensive."

Payments. In the US right now, iOS apps can link straight to web checkout with no entitlement and no commission, because of the Epic injunction. Apple's guidelines carve the US out three separate times. But the Ninth Circuit ruled in December 2025 that the zero-commission remedy was an abuse of discretion, and the Supreme Court took the case on June 30, 2026 with argument expected this fall. The 0% US rate is a temporary legal artifact.

Outside the US it's already settled and it's bad: Patreon's own help center, updated May 6, 2026, states that for international fans "you cannot opt out of Apple's 30% fee." Patreon marks international iOS prices up 42.86% to keep creator earnings flat, and Apple gave it a hard November 2026 deadline. Any platform selling memberships in-app internationally will eventually get the same letter.

The safe harbor is Apple's 3.1.3(f): a free companion app to a paid web service needs no in-app purchase, provided there is no purchasing inside the app and no calls to action for purchase outside it. Silent app, web-only checkout. That's the architecture the plan below uses.

2. Distribution, and the number is brutal

Skool pays a 40% recurring-for-life affiliate commission with a 60-day cookie. On the $99 plan that's $39.60 a month to the referrer, forever. Layered on top: Alex Hormozi's 2023 lead investment (which he called the largest personal investment of his life), his migration of the largest audience in business education onto the platform, and the Skool Games monthly cash-prize leaderboard. Skool did not win on product. It won by making thousands of creators financially invested in recruiting other creators.

Now do the arithmetic on a cheaper competitor. At $39/month with 6% monthly churn: average lifetime 16.7 months, gross LTV about $650, contribution margin maybe $520. Self-serve SaaS CAC in 2026 runs $420 to $702. That's an LTV to CAC ratio of about 1.2 to 1. It does not work. Healthy is 3:1.

Cutting price and needing affiliates are directly contradictory. Skool can afford 40% forever because it charges $99, not $39.

What is emphatically not a problem: infrastructure

Per community, per month, on your existing Cloudflare stack, at 2026 list prices:

Live video is 68% of the 1,000-member bill and 74% of the 10,000-member bill. Everything else is a rounding error. At $89/month, infrastructure gross margin is 85% at 100 members and 92% at 1,000. Even undercutting Circle to $39 leaves 82% margin at 1,000 members.

And here's the one genuinely striking arbitrage in this whole research: buying chat infrastructure costs $499/month (Stream, 10k MAU) or $399 to $499/month (Sendbird, 5k MAU). Running it yourself on Cloudflare Durable Objects costs about $3.47/month at 10,000 members. Durable Objects bill outgoing WebSocket messages at zero and incoming at a 20:1 ratio, which is exactly the right shape for community fan-out. That's a 100x input-cost gap, and it sits on the stack you already run.

(Two cautions on Durable Objects, both from primary 2026 sources: a documented case of $34,000 billed in 8 days from a runaway alarm loop with zero users and no platform warning, and Wire's July 2026 engineering post-mortem about placement being fixed at creation time forever. Set hard billing alarms.)

---

Recommended plan

You said you want to sell this to other creators, so this plan is built as a SaaS. But it's built in the shape the evidence supports, not the shape that loses.

The positioning

Do not sell "cheaper Circle." Sell "the community platform you can leave, that never takes a percentage."

Three things make that a real wedge instead of a price cut:

1. 0% transaction fees, permanently, in writing. Circle takes 2%, Skool takes 10% at the low tier, Heartbeat takes 5% on top of Stripe. At $10,000/month in member revenue, Circle's Professional plan costs $89 + $200 = $289/month. A flat $79 with 0% is genuinely cheaper for anyone succeeding, and more expensive for anyone failing, which is the correct incentive alignment and a great story.
2. Full portability guarantee. One-click export of members with payment history, plan, tenure, and tags. This is the thing every founder in the research said almost never exists, and it's the thing that would make someone trust a small vendor. Turn the biggest objection to buying from an unknown ("what if you disappear?") into the headline feature.
3. One vertical you can actually reach. The evidence is unambiguous that a solo founder cannot out-market Skool's affiliate army horizontally. The only viable acquisition channel is an audience you already own. Every existing platform's design language (dark modes, gamification, leaderboards, Discord density) actively repels the 55-plus health and wellness audience, and you have both credibility and existing reach there.

The architecture, and why

• One aggregator app. A single "picker" binary under your own Apple and Google accounts, hosting every customer's community. Apple's 4.2.6 explicitly blesses this. One developer account, one review queue, one SDK migration a year, 3 to 6 weeks of annual compliance instead of 40 to 120 days.
• Web-only checkout, always. No purchase flows and no purchase calls-to-action inside the app, which lands squarely in Apple's 3.1.3(f) safe harbor and avoids the 30% international cut entirely. Circle Plus's in-app payments are a liability, not a feature.
• Branded apps as a $500/month enterprise upsell you sell at most five of. Priced to cover the human labor honestly, using the customer's own developer accounts (the Honeycommb and Mighty Pro model). Never as a mass-market feature.
• Cloudflare all the way down. Workers, D1, R2, Durable Objects for chat and presence. This is the arbitrage and it's your existing stack.
• Postgres-shaped multi-tenancy with row-level tenant isolation enforced at the database boundary, not in application code, and applied to background jobs too (that's where tenant context most often gets dropped).
• Expo for mobile. Cloud builds mean no Mac required, over-the-air updates ship JS fixes without an App Store review cycle, and the same TypeScript covers web and mobile. Flutter genuinely wins the benchmarks, but for a feed-and-forms app that doesn't matter and you'd be learning Dart.
• Do not build on ActivityPub, Matrix, or the AT Protocol. The one well-built open-source community platform on AT Protocol (Colibri) is stuck in beta specifically because the protocol has no permissioned-data story yet, and private paywalled content is the entire product. Boring Postgres.

Phases

Phase 0, weeks 1 to 8. Prove the buyer exists before writing the platform.

Do not build first. Run one real paid community yourself on cheap existing software: self-hosted Discourse (~$30/month, and it ships spaces, chat, events, gamification, and working Stripe subscriptions in core) wrapped with Natively ($19/month) for a phone app. That's a Circle-class private community for under $50/month with no code written, running this month.

Then talk to 20 people in the target vertical who currently pay Circle, Skool, or Mighty. Two questions, verbatim from the research: would you leave your current platform for one that takes 0% and lets you export everything including payment history? and what would have to be true for you to actually move? If fewer than 5 say yes convincingly, stop here, and you'll have spent about $200 and eight weeks instead of a year.

Phase 1, months 3 to 6. Web MVP, one vertical, no mobile.

Auth and tenancy, spaces, posts, threaded comments, reactions, member profiles and directory, roles and permissions, Stripe Connect with 0% platform fee, email notifications and a weekly digest, basic moderation, and admin export with full payment history. Ship to 10 design partners at a founding price. No chat, no courses, no live video yet. This is 10 to 16 weeks of focused work.

Phase 2, months 7 to 10. Chat and the aggregator app.

Real-time chat on Durable Objects (this is the arbitrage, and it's 4 to 8 weeks). Push notifications. The single Expo aggregator app on both stores. Web-only checkout maintained strictly.

Phase 3, months 11 to 14. Depth.

Courses with drip and progress. Events with RSVPs. Gamification. Automations. This is the tier that lets you charge $79 to $99 instead of $39.

Deliberately not building, at least for a year: live streaming and live rooms (74% of infra cost and the hardest ops), an API, custom SSO, AI agents, and per-customer branded apps.

Pricing

$79 is the anchor, not $39. The research is clear that $99 is the accepted market reference price and that discounting to $39 destroys the LTV-to-CAC math while buying almost no differentiation. Sell 0% fees and portability, not the sticker price.

Kill criteria, agreed in advance

Stop and shut it down if any of these are true:

• Fewer than 5 of 20 Phase 0 interviews say they'd switch.
• Fewer than 10 paying customers by month 9.
• Monthly customer churn above 8% for three consecutive months.
• Customer acquisition cost above $300 with no organic channel working.
• Apple or Google rejects the aggregator app twice for reasons you can't engineer around.

Honest odds

5 to 8% for "cheaper Circle with branded apps for everyone." The two halves fight each other: cheap means no margin to fund the per-customer app store labor that branded apps require, and that labor doesn't scale with software.

25 to 35% for the plan above: one vertical you can reach, one compliant app, web-only payments, $79 not $39, branded apps as a rare upsell. That's a real if modest business, and the infrastructure math genuinely supports it at 85%+ gross margin.

The single most likely cause of failure is not Apple rejecting anything. It's operational drowning, and right behind it, the impossibility of out-marketing a 40%-forever affiliate program without an audience you already own. Which is exactly why the vertical, and Phase 0, are load-bearing rather than optional.

The strongest argument against all of it

Worth stating plainly, because it's good.

Your buyer is not choosing between Circle at $89 and you at $79. They're choosing between paying anything at all and using the free community layer inside a platform their audience already lives in. YouTube Communities shipped and kept expanding through July 2026. Patreon rebuilt its mobile app from scratch, shipped chats, Shops, and short-form posts, and redesigned its homepage into a recommendation feed. Substack now ships publication chat, subscriber DMs, threaded comments, live video, and Notes as a discovery network, at 10% and no monthly fee.

And Geneva is the lesson that should sting: real users, real brand affinity, a $17.5M acquisition, zero revenue, shut down September 2025. The failure mode in this category is not "nobody wants it." It's "people want it, love it, and won't pay for it, while the ones who will pay are locked in somewhere else."

---

Confidence and gaps

HIGH: all pricing, plan limits, add-on costs, and transaction fees (read from vendor pricing pages August 1, 2026); Circle's tier history (Wayback-verified); Apple and Google guideline text (fetched verbatim); GitHub license, star, and activity data (queried live); Cloudflare, Supabase, Convex, LiveKit, Mux, and SES pricing (primary docs); Frond's shutdown and Geneva's acquisition and closure.

MEDIUM: Circle Plus and Mighty Pro actual contract prices (both sales-gated, public estimates range from $360/month to $2,500/month, and the frequently-cited ~$30,000/year figure for Circle Plus could not be corroborated against any primary source, so do not build a business case on it). Whop's exact fee stack. EzyCourse's white-label app tier threshold. Skool's user and revenue figures (third-party estimates disagree by 2x). Churn benchmarks, which are mostly vendor-adjacent content marketing rather than audited data.

LOW: the build-effort estimate. There is no verified public example of a solo founder shipping and monetizing a full community platform with native apps, so the 40-to-60-week parity figure is a bottom-up component estimate, not an observed outcome. Also low: how Circle Plus branded apps handle developer account ownership and Apple's IAP cut, which Circle does not disclose and which should be asked of their sales team directly if this proceeds.

Not researched: Reddit blocked direct fetching, so Reddit quotes come from cached indexes with real URLs but incomplete upvote data. X is login-gated and effectively unsearchable; no creator complaint threads were found there and none were invented. G2 blocked extraction, so star distributions are absent.