How to Run a Paid Training Community Profitably

Run a paid training community by pricing from both sides: set the member price from what your niche pays, then subtract the fee stack on every payment, your metered delivery allowance, and the facilitator hours each tier consumes.

The published advice on this is good and it is all demand-side. It tells you to anchor on outcomes, benchmark your niche, cap yourself at three tiers, and bill quarterly. What it does not do is finish the arithmetic. Every guide reaches the sentence "add up your costs" and moves on, which leaves the operator holding the only question that decides whether the thing survives: at this price, with this tier mix, what is left over per member, and how many members can the plan actually serve?

A training community is where that omission bites hardest. A discussion community costs the same to serve whether a member posts once or a hundred times. A training community does not. Live calls have participant caps, voice practice is metered by the minute, and coaching hours scale one-to-one with headcount. The unit economics are closer to a services business than to a forum, and they need to be modelled that way.

What does it actually cost to serve one member?

Three layers: a fee stack taken from every payment, a metered delivery allowance consumed per session, and facilitator time that scales with your highest tier. Only the first is a percentage. The other two are capacity, and capacity runs out before revenue does.

Separating these matters because they fail differently. Fees shave a predictable slice off every payment. Capacity does not shave anything until it hits a wall, and then your core promise stops working for everyone at once.

What does the platform actually keep from each payment?

A platform take rate plus a payment processor fee. Published platform rates range from 0.5% to 10% depending on plan tier, and Stripe lists 2.9% + $0.30 per successful domestic card transaction. Whether both apply to the same charge depends on the platform, and it is worth asking directly.

The numbers below are taken from each company's own public pricing page as of 17 August 2026. Rates move, so re-read them before you model anything.

Two details in that table do more damage than the headline percentages. The first is the flat per-charge component, which is a much larger share of a $29 payment than a $249 one. The second is the entry-tier rate: a 10% take on a starter plan is roughly double what the same platform charges one tier up, which means the cheapest plan is only cheap while your revenue is small.

Why does the premium tier have the worst margin?

Because coaching hours scale one-to-one with members while community access does not. A $249 coached tier can return a lower percentage margin than a $29 async tier, even though it returns far more money per member. Percentage margin and absolute contribution point in opposite directions here.

The table below models three tiers on Voxento's Business plan fee of 4.9% + 30 cents per charge, with facilitator time costed at $60 per hour. The live call in the middle tier is one shared hour spread across 25 members. The 1:1 time in the top tier is 45 minutes per member per month.

Read the last two columns together and the useful conclusion appears. The coached tier is the best tier to sell — $191.50 a month beats $27.28 by a wide margin. It is also the only tier whose cost grows with every sale, and it consumes twice the metered voice allowance of the tier below it. Sell enough of your best tier and you hit a ceiling that has nothing to do with money.

What caps your community before your seat limit does?

Metered capacity. Plan pages advertise a member seat count, but the binding constraint is usually the live-video participant cap or the included voice minutes. Both typically run out at a fraction of the seats you are entitled to fill.

Voxento's published add-ons make the arithmetic checkable. The Hume add-on on the Business tier is $77 a month and includes 140 script minutes and 200 agent minutes. The Live Kit add-on is $50 a month for unlimited video calls with up to 30 participants. The Business plan itself covers up to 50 team members. Those three numbers do not describe the same size of community.

• An 8-minute roleplay drill against a 200-minute monthly allowance is 25 drills a month for the whole community, not per member.
• Spread across 40 members, that is roughly one drill each every two and a half months.
• Promising every member one drill a week means 40 x 4 x 8 = 1,280 minutes a month, which is 6.4 times the included allowance.
• A 30-participant video cap means a 40-member community cannot hold a single all-hands call. You are running two sittings, which doubles the facilitator hour, not the cost of the seat.
• Overage pricing beyond the included minutes is not published, so treat the allowance as a hard ceiling until someone tells you the per-minute rate.

Now put the tier mix back in. Twenty-five members on the middle tier at 16 minutes each is 400 minutes. Five on the coached tier at 32 minutes each is another 160. That is 560 minutes against an allowance of 200 — nearly three times over, from a community of thirty people. The premium tier you were told to sell harder is the one that breaks the allowance first, and nothing on the pricing page warns you.

This is the same constraint from the delivery side that shows up on the design side: voice practice is metered and capped by concurrency in a way text practice is not. If you have not read why modality changes what a training programme costs to run, that post covers the scheduling half of the problem this one costs out.

Does quarterly billing do anything besides reduce churn?

Yes. It cuts the flat per-charge portion of your fee stack by two-thirds. Twelve charges a year become four, and each avoided charge saves the fixed amount on both the platform fee and the processor fee. Nobody mentions this because the retention argument is louder.

The retention case is well made elsewhere and it is the bigger effect. Circle's survey of 55 community builders found operators dropping monthly billing entirely because it did not create enough commitment. The billing-cost case is smaller but it is free money, and it points the same direction.

At a 30-cent flat component, moving a member from monthly to quarterly saves 8 charges a year, or $2.40. Across 200 members that is $480 a year for a change that also improves retention. If a separate processor fee with its own flat component applies to the same charge, double it. That is not a large number, but it costs nothing to capture and it compounds with the churn effect rather than trading against it.

The contribution-margin worksheet

Ten inputs, six formulas, and one capacity check. Rebuild it in a spreadsheet in about ten minutes. It answers two questions the demand-side guides cannot: what one member is worth after everything, and how many members your plan can actually serve.

Inputs to gather first

1. Member price per tier, and the billing frequency for each.
2. Platform percentage take rate, from the platform's own pricing page.
3. Platform flat fee per charge.
4. Processor percentage and flat fee, if these apply separately. Confirm this rather than assuming.
5. Fixed monthly platform cost: base plan plus every add-on you actually need.
6. Included metered allowance per month: voice or session minutes, live hours, participant caps.
7. Minutes consumed per member per month, by tier.
8. Facilitator hours per month, split into shared hours and per-member hours.
9. Your real hourly rate for facilitator time, including your own.
10. Expected member count per tier.

The six formulas

• Fees per member = (price x platform rate) + platform flat fee + (price x processor rate) + processor flat fee
• Net revenue per member = price - fees per member
• Facilitator cost per member = (shared hours x hourly rate / members in tier) + (per-member hours x hourly rate)
• Contribution per member = net revenue per member - facilitator cost per member
• Contribution margin = contribution per member / price
• Monthly surplus = sum of (contribution per member x members) across all tiers - fixed monthly platform cost

Fixed platform cost sits outside the per-member calculation on purpose. It is the number your total contribution has to clear before anything reaches you, and keeping it separate stops it from being quietly amortised into a per-member figure that looks healthier than it is.

Worked example: 40 members on one tier

The capacity check

Run this second, and run it before you publish the price rather than after you have sold the tier.

1. Minutes needed = sum across tiers of (members in tier x minutes per member per month).
2. Divide by your included allowance. Anything above 1.0 means you are relying on overage pricing you may not have confirmed.
3. Compare your largest single-session attendance to the participant cap. Above the cap means repeat sittings, and each repeat is another facilitator hour.
4. Re-run both at the member count you are targeting in twelve months, not the one you have today.

If step 2 comes out above 1.0 you have three options and only three: raise the price to fund the overage, cut minutes per member, or move the practice to a format that is not metered. Deciding which before launch is a pricing decision. Deciding it after launch is a refund conversation.

What should you confirm before you publish a price?

Six things, all answerable in an afternoon from public pages and one support email. Every one of them changes the model materially, and none of them are visible on a pricing page's headline.

• Does the platform take rate sit on top of a separate processor fee, or does it include it?
• What is the per-unit price once you exceed the included metered allowance?
• Does the processor return its fee when you issue a refund? Budget a refund reserve on the pessimistic answer.
• What does a disputed charge cost? Stripe publishes a $15 dispute received fee, and a $15 fee for manually countering one, refunded to you on disputes you win.
• Does the take rate change between plan tiers, and at what revenue does upgrading pay for itself?
• Is the participant cap per call or per account, and does it apply to recorded sessions too?

The last one catches people. A cap that applies per call is a scheduling problem. A cap that applies per account is a product ceiling, and it decides how large your community can get before the format has to change.

Once the model clears, the rest of the operating question is content design rather than arithmetic: what members practise, and how you tell whether it worked. The AI roleplay training pillar covers when practice beats passive content, scoring a roleplay covers making the assessment defensible, and the objection scenario library is a set of drills you can run without writing them yourself. If you want to check the fee and allowance figures used here against the current plans, they are on the Voxento pricing page, and you can try a live voice conversation to see how long a real drill actually runs before you budget minutes for it.