Founders Pro at $19/mo — Cost + Pricing Rationale

Status: research-only. This document does NOT constitute legal or tax advice. Before finalizing the price or committing the SKU to Stripe, consult a CPA licensed in Pennsylvania on COGS treatment, revenue recognition, and PA sales-tax nexus. Last updated: 2026-04-23. Vendor prices are from the sources listed — verify freshness before any financial commitment.

TL;DR

Founders cohort is capped at 250 paying users (per docs/marketing/pricing.md §open questions — the 250 framing predates the current Founders Promo and carries forward as the cohort ceiling). Unit economics below are scoped to that cap, not to unbounded user counts.

At 10 paying users, $19/mo is below cost. At roughly 40–50 paying users, $19/mo reaches break-even. At 100 users it generates ~$13.50 net margin per user. At the 250-user cap, it generates ~$16–17 margin per user (~$4,200/mo gross margin on the cohort).

The "at-cost" framing is only accurate for the first 14–25 users. From there up to the 250 cap, $19 is a Founders discount below standard Pro pricing, not literally at-cost. If the public narrative promises "at cost," the CPA should weigh in on whether that promise has accounting or contractual implications once the cohort clears ~25 users.

A more honest public framing for the full 250-seat cohort: "Founders pricing: locked at $19/mo — below our standard Pro sticker because you bought in early." That holds at 10 users (below cost, accepted subsidy) and at 250 users (below-sticker, not below-cost) without lying about either end.

Users 251+ pay standard Pro pricing (different economics, different analysis — see docs/marketing/pricing.md). This doc only analyzes the Founders cohort.

1. Cost Structure

1.1 Fixed Costs (amortized across all paying users)

These costs are incurred regardless of user count. They do not scale with individual Pro users.

Values below refreshed 2026-04-24 from docs/business/bookkeeper/2026-04-reconciliation.md (YTD 2026 through April ledger). Four items confirmed from the ledger; three marked ⚠️ still pending dashboard verification.

What remains placeholder: only the Sentry paid-tier (a forward decision, not a dashboard check). All three dashboard items (Alpaca / Dreamhost / Dyn) confirmed 2026-04-23 from vendor dashboards.

1.2 Variable Costs (scale with user count or revenue)

These costs grow as the Founders Pro cohort grows.

Heroku compute (Raptor backend)

Current stack: Heroku dynos + Heroku Postgres + Heroku Key-Value Store (Redis), per docs/finance/vendor-conventions.md.

Per-user Heroku share: Heroku compute is a shared resource. With 10 Pro users, each user "consumes" a proportional share of the fixed dyno + database bill. At 100 users, the compute doesn't scale linearly — it's a semi-fixed staircase (upgrade dynos or Postgres tiers when you hit capacity). Estimate: 1 Standard-1x web dyno supports ~50–200 concurrent light-use sessions before needing a second dyno.

Alpaca Market Data API

Architecture: one server-side Alpaca account shared across all users — not per-user. Per docs/architecture/mbt-paper-trading-engine.md §7 and docs/architecture/multi-tenant-alpaca.md §3.

Source: alpaca.markets/data — Algo Trader Plus cited at $90.75/mo across community sources; verify current price on alpaca.markets/data directly, as Alpaca has restructured data offerings.

Per-user Alpaca share: The $90.75/mo is a fixed cost once you have any Pro users at all. It amortizes rapidly: at 10 users = $9.90/user; at 50 = $1.98/user; at 100 = $0.99/user.

Anthropic API (AI trade proposals)

Pro tier: 500 AI proposals/mo per user. Per docs/marketing/pricing.md.

Estimate per proposal (Haiku 4.5, short prompts ~500 tokens in / ~300 out): - Input: 500 tokens × $0.001/1k = $0.0005 - Output: 300 tokens × $0.005/1k = $0.0015 - Cost per proposal ≈ $0.002 (Claude Haiku 4.5)

If Sonnet is used instead (better quality, higher cost): - Input: $0.003/1k = $0.0015 + Output: $0.015/1k = $0.0045 - Cost per proposal ≈ $0.006 (Claude Sonnet 4.6)

Per platform.claude.com/docs/en/about-claude/pricing (as reported by multiple 2026 sources).

Flag: The actual model used is not pinned in the architecture docs. Identify the model before committing. For this analysis, Haiku at typical usage ($0.20/user/mo) is used as the base case; Sonnet at max usage ($3.00/user/mo) is the stress case.

Stripe fees

Per transaction: 2.9% + $0.30 (US standard). Source: stripe.com/pricing.

On a $19 charge: ($19 × 0.029) + $0.30 = $0.55 + $0.30 = $0.85/transaction

As a fraction of revenue: $0.85 / $19 = 4.5% effective rate on a $19 charge.

Note: Stripe's fixed $0.30 per-transaction makes the effective rate higher on lower monthly charges — this is a reason low price points ($9–$14) are proportionally more expensive to process than higher ones ($29).

1.3 Semi-Variable Costs

2. Unit Economics at Three User-Count Tiers

Base assumptions: - Heroku: Standard-0 Postgres ($50) + 2 Standard-1x dynos ($50) + Redis Mini ($3) = $103/mo - Alpaca: $90.75/mo (Algo Trader Plus, one shared account) - Anthropic: $0.20/user/mo (Haiku, 100 proposals typical) - Stripe: $0.85/transaction/mo per user - Fixed overhead (Workspace + Dyn + Cloudflare + Dreamhost): $40/mo (lean, attorneys quiet) - Attorney/CPA amortized: $150/mo (when active, distribute across all users)

10 paying users

At 10 users: below cost by ~$3.25/user/mo. Total monthly loss: ~$32.

100 paying users

At 100 users, assume Heroku dyno upgrade to 2 Standard-2x ($100) + Standard-0 Postgres ($50) + Redis Mini ($3) = $153/mo. Alpaca still one account ($90.75/mo).

At 100 users: healthy margin of ~$13.53/user/mo. The at-cost framing does not hold; this is genuinely profitable. If "at-cost" is the brand promise, $19 is defensible here only as a "we're not gouging you" posture, not literal cost-recovery.

250 paying users — the Founders cap

At 250 users (the full Founders cohort cap), infrastructure assumptions: 1–2 Standard-1x dynos ($50–$100) + Postgres Essential or small upgrade ($20–$50) + Redis Mini ($15) = ~$85–$165/mo. Alpaca: standard Algo Trader Plus tier still supports this scale (well under multi-tenant data limits) — no upgrade required at 250 users. Anthropic API scales ~linearly with user count.

At 250 users (the cap): $19 yields ~$4,000/mo in gross margin for the cohort. This is genuinely profitable SaaS — not "at cost." The honest framing at this user count is a Founders discount below standard Pro pricing, not a cost-recovery claim.

The transition point: sub-25 users it's an accepted subsidy ("we're eating cost to seed the cohort"); 25–250 users it's a discount offer ("you paid less than standard Pro because you bought in early"). The price doesn't change — the honest narrative does.

Break-even user count at $19/mo

Break-even is where total revenue = total costs.

Using base assumptions (fixed + Heroku + Alpaca as stated):

Break-even range: 14–22 users, depending on fixed-cost run rate.

3. Sensitivity Analysis

S1: Alpaca data cost 2× ($198/mo instead of $99)

New break-even: ($292 fixed + $99 added) / $17.95 = ~22 users (lean) or ~27 users (with attorneys). Margin at 100 users drops from $13.53 to $12.54/user (-7%). Manageable.

S2: Heroku dyno share 2× (need to double compute earlier)

Adds ~$50–$100/mo. Break-even shifts from ~14–22 to ~19–28 users. At 100 users, margin drops by $0.50–$1.00/user. Not material at scale.

S3: Anthropic API 10× (Sonnet at max 500 proposals/user)

At $3.00/user/mo instead of $0.20:

• Break-even with $242 fixed: $242 / ($19 - $3.85) = ~16 users (vs. 14 base). At 100 users, margin drops from $13.53 to $10.73/user. Still profitable.
• The real risk: if Raxx leans heavily into LLM features and per-user token consumption grows beyond 500 proposals (e.g., backtesting commentary, trade explanations), the $3/user budget assumption breaks down. Flag: define Anthropic usage quotas per tier before pricing is finalized.

S4: Price floor comparison

The $24 price point is the first where 10 users is cash-flow positive. At $19, the product operates at a loss until ~14–22 paying Founders are live.

4. Recommendation

This section presents the research findings. It does NOT constitute a recommendation to charge any specific price. Take this to a CPA before committing.

Research finding: $19/mo is viable as a Founders Pro price with conditions:

1. User-count threshold. Break-even is ~14–22 users. If the Founders cohort converts fewer than 15 users to paid, $19 produces a monthly operating loss. The product cannot sustain $19 indefinitely at single-digit user counts.

2. Time-bounded framing is correct. The retired "$19 locked forever" offer (docs/marketing/pricing.md §open question #1, now retired per issue #204) was riskier than it appeared — at perpetual pricing, $19 is a 34% discount to Pro ($29) with no path to recover the gap as infrastructure matures. The current framing (post-trial conversion, not locked forever) is materially safer.

3. The "at-cost" narrative is accurate only at year-one scale (10–25 users), then stops being accurate. At 100 users, $19 produces ~$1,350/mo gross margin. At 250 users (the Founders cap), ~$4,000/mo gross margin. That's not "at-cost" — it's profitable SaaS. Recommended correction: retire the "at-cost" language for public-facing copy past the first ~25 users. A more honest + defensible framing for the full 250-seat cohort is "Founders pricing: locked at $19/mo — below our standard Pro sticker because you bought in early." That's true at both ends of the cohort without lying about either.

If the at-cost framing IS kept as a public promise, the CPA should weigh in on whether that has accounting or contractual implications when the cohort clears the break-even threshold.

4. $24 is the price point where 10 users breaks even. If the Founders cohort is expected to be 20–50 users (year-one realistic), $19 and $24 both reach break-even, but $24 carries more buffer for cost surprises (Heroku upgrades, Alpaca tier increases, Anthropic model choices).

5. $19 vs. $29 Pro cannibalization. Per docs/marketing/pricing-v2.md §explicit call: "a move down to $19 would undercut our own Founders promo and signal we don't believe in the ceiling." That framing was written when $19 was the promo price and $29 was standard Pro. Under the current plan (Founders trial → $19 at-cost → eventual $29 standard), the $10 delta is the incentive for Founders to convert rather than churn. That delta is reasonable.

Alternative price points:

Single-price research finding: $19 is numerically defensible as an at-cost price given year-one user-count expectations. If the Founders cohort reaches 20+ paying users within the first 3 months of the paid transition, $19 covers costs. If it doesn't, $24 is the safer floor with the same "not standard-Pro-priced" positioning.

Confidence: Moderate. The largest uncertainty is the Alpaca data tier (verify current pricing; architecture docs flag this as an open question) and the Anthropic per-user token consumption (no usage data yet).

5. CPA Questions

The following questions require a CPA licensed in Pennsylvania before the Founders Pro SKU is committed to Stripe or communicated publicly.

1. COGS vs. operating expense for SaaS infra in a PA LLC (disregarded entity or S-Corp). Heroku, Alpaca, and Anthropic API costs are arguably COGS on a gross- margin view of the business. For a Schedule C (current posture) vs. Form 1120-S (post-S-Corp election), how should these be classified? Does COGS treatment change what "at-cost" means for tax purposes?

2. Revenue recognition for the Founders cohort. If Founders pay $19/mo on a monthly subscription, revenue is recognized monthly (cash or accrual basis?). Is there any deferred revenue concern if a Founder pre-pays an annual commitment? Does PA LLC status change the treatment?

3. Is "at-cost" pricing a taxable event or gift/promotional concern? If the product is genuinely priced below market rate (standard Pro = $29, Founders rate = $19), is the $10 delta a promotional discount (taxable revenue shortfall only), or does it create any charitable-contribution-like accounting treatment at the entity level? Confirm "at-cost" is not "below-cost" once the business exceeds ~25 users.

4. PA sales-tax nexus on SaaS subscriptions. Pennsylvania has a complex SaaS tax treatment — some digital services are taxable. At what revenue threshold does Raxx need to collect and remit PA sales tax? Does Kristerpher's part-year presence in California (4 months/year) create CA nexus for the business, and if so, does CA's SaaS tax posture apply?

5. Home-office deduction interaction. docs/finance/chart-of-accounts.md defers home-office computation to the CPA. If a home-office deduction is claimed, does that affect how Heroku/Alpaca/Anthropic costs are allocated? (Deferred by bookkeeper convention — confirm the CPA handles this at tax time.)

6. Source Citations

8. Annualized view (added 2026-04-24)

Per-month figures multiplied by 12 for the CPA conversation, investor-adjacent framing, and forward-planning sanity checks. All at the locked $19/mo Founders price.

Revenue + margin annualized across Founders cohort tiers

What the annualized cap figure means

$57,000/year gross revenue at full Founders capacity, with ~$48,900/year gross margin after Stripe + Heroku + Alpaca + Anthropic + fixed allocation. That's not business-sustaining income on its own — but it's meaningful validation revenue and covers infrastructure + meaningful operating headroom while the standard-Pro tier ramps alongside.

What the annualized loss at 10 users means

At 10 paying Founders, we lose ~$1,750/year on the cohort. That's the "accepted subsidy" cost of the pre-break-even period — the price of seeding the early cohort. If the cohort plateaus at 10 users, that loss is the real annual burn on the Founders line. If it grows to 100, the loss flips to a $16k gain within that same year.

What's NOT modeled here

• Users 251+ — these pay standard Pro (different economics, see docs/marketing/pricing.md). Annualized Pro revenue is computed separately.
• Revenue recognition across the trial window — a Founder who joins in month 1 but doesn't convert to paid until month 4 generates $0 during the trial. The annualized table above assumes steady-state paid cohort, not trial→paid ramp.
• Churn — the annualized figures assume every paying Founder stays paid for 12 months. Real-world churn should be modeled per cohort once we have data.
• Income-tax treatment — these are gross-margin figures, pre-federal/state income tax, pre-self-employment tax, pre-S-Corp-owner-comp. The CPA handles those layers.

Relationship to standard Pro economics

Standard Pro at $29/mo (per docs/marketing/pricing.md) generates $348/year per paying user. If the standard cohort reaches 100 paid users year-one (plausible if the waitlist delivers), that's $34,800/year — comparable to the full-capacity Founders cohort. Standard Pro is the scalable revenue line; Founders is the validation cohort. Don't conflate the two when planning.

9. Dashboard verification — COMPLETED 2026-04-23

All three values returned from vendor dashboards. Applied to the cost tables above. Summary below for audit trail; see individual table rows for live numbers.

Confirmed values

Reconciliation flag for bookkeeper

Dreamhost YTD vs forward rate mismatch: the bookkeeper ledger shows 2026-01-29, 02-28, 03-29 PayPal charges all at $5.00, but the dashboard shows the PHP Extended Support 2+ plan currently at $10/mo — implying a rate change occurred at or before 2026-05-04. Additionally, the Shared Unlimited $155.88 annual charge will hit 2026-10-29; not yet in YTD data. Next bookkeeper close should catch both the April PHP charge (at whichever rate) and forecast the October Shared Unlimited annual renewal. Flag added to docs/business/bookkeeper/ for the next reconciliation pass.

Original to-do items (retained for audit)

The three checks originally listed here — Alpaca dashboard, Dreamhost panel, Dyn services — were completed 2026-04-23 by Kristerpher. Screenshots retained in the session transcript. Values above are authoritative.

9.1 Alpaca data tier — confirm current plan + monthly cost

• Go to: https://app.alpaca.markets/brokerage/dashboard (or https://broker-app.alpaca.markets depending on your account type)
• Look for: Settings → Plan / Subscription → current tier + monthly cost
• Expected: "Algo Trader Plus" at $90.75/mo (IEX quotes), OR a free IEX-only tier, OR something else
• Report back: tier name + monthly $ amount. If you're on free IEX-only, we revise this doc's cost assumptions downward meaningfully (Alpaca goes from $90.75/mo to $0 in the fixed-costs table)

9.2 Dreamhost — confirm exact plan name

• Go to: https://panel.dreamhost.com → Billing → Active Services
• Look for: the plan name attached to moosequest.net hosting (we know the monthly charge is $5; plan name is what's missing)
• Expected: something like "PHP Extended Support 2+ Monthly Plan" (per the invoice PDFs) or similar
• Report back: exact plan name so the doc can reflect it. No pricing change expected — just nomenclature.

9.3 Oracle Dyn — confirm plan name + shutdown timing

• Go to: https://account.dyn.com → Services
• Look for: the DNS plan name currently active + any shutdown-notice banners
• Expected: "Dyn Standard DNS" or similar; Oracle announced Dyn retirement 2019 with May 2023 hard date (currently still running on retail). #216 tracks the migration.
• Report back: current plan name + whether there's a new shutdown-date banner. If shutdown date is now imminent, escalate #216 priority.

Once you hand me values for any/all three, I update the doc in a follow-up commit + drop the ⚠️ flags. None of these change the break-even analysis meaningfully — they refine the line items from placeholders to confirmed.

Before finalizing the price, consult a CPA on COGS classification, revenue recognition for a monetized Founders cohort, PA + CA sales-tax nexus, and whether the "at-cost" framing holds accounting water at scale. Recommend: CPA licensed in Pennsylvania with SaaS / digital-goods experience. Secondary: tax attorney if the PA SaaS sales-tax nexus question is not clearcut.