Date: 2026-04-21
Author: TITAN / SCOUT Research Arm
Status: Internal Working Document — Not for Distribution
Word Count Target: 3,000–4,000 words
---
This memo exists to make the revenue model concrete, defensible, and ethically aligned before we acquire a single paying user. The decisions documented here are load-bearing. They will be tested by investors, regulators, competing product teams, and most importantly by users who have placed real emotional trust in this product. We owe all of them the same level of rigor.
Where we have data, we cite it. Where we are estimating, we say so. Where we are making a values choice rather than an optimization choice, we name it as such.
---
Silent Infinity occupies an unusual position in the consumer software market: it is a product whose core value proposition requires users to be emotionally vulnerable. People come here after a panic attack, after a breakup, after 3am when the ordinary scaffolding of their life has collapsed. That context creates an obligation that most SaaS products never have to reckon with. The revenue model is not a peripheral operational decision — it is a moral architecture.
Adrian Slywotzky's Profit Patterns (2003) describes three dominant monetization archetypes relevant to digital services. The Ad Model generates revenue by selling user attention to third parties. The Engagement Model generates revenue by maximizing time-on-platform, session length, and return frequency. The Customer Relationship Model generates revenue by providing sustained value for which users pay directly.
We explicitly choose the Customer Relationship Model and explicitly reject the other two. This is not a branding decision. It is structural.
The Ad Model is incompatible with Silent Infinity for a simple reason: the mirror cannot have advertisers. A product that asks users to examine their inner life without judgment cannot simultaneously be monetizing that inner life to a wellness supplement brand, a meditation app competitor, or a pharmaceutical advertiser. The moment ad revenue enters the model, every design decision about what content surfaces, how long a session lasts, and what emotional state the user exits in becomes contaminated by a second master. We will not build that product.
The Engagement Model is more insidious because it masquerades as caring about the user. Duolingo's streak mechanics, Calm's daily reminder push notifications, the red badge on every app icon — these are not features. They are operant conditioning loops engineered to maximize MAU/DAU ratios for investor decks. For most consumer software this is merely annoying. For a mental health-adjacent product, it is corrupt. A user who feels guilty for missing three days of "their practice" has been harmed by our product design, not helped by it. Revenue must not rise when a user spirals. If our P&L improves when someone's anxiety worsens and they're using Silent Infinity more, we have built the wrong thing. We have built an anxiety product, not a calm product.
The subscription model aligns incentives cleanly. The user pays a fixed amount per month. We earn that payment by being worth paying for. If we are not worth paying for, they cancel. There is no mechanism by which making them feel worse increases our revenue. There is no mechanism by which designing compulsion loops increases our revenue versus simply building a product they find genuinely valuable. This is not naivete — it is structural incentive alignment, and it is the foundation on which every other product decision should be tested.
The structural commitment, stated plainly: Silent Infinity's revenue does not rise when a user spirals. If a design decision would increase engagement but increase distress, it is rejected. This is not a policy we vote on. It is a load-bearing constraint on the business model.
---
Pricing tiers exist to capture value across different willingness-to-pay segments while keeping the core product accessible to everyone. The free tier is not a trial. It is a permanent, dignified product. The paid tiers offer genuine additional capability, not the removal of artificial restrictions.
| Feature | Free | Plus ($9/mo) | Pro ($19/mo) | Org ($99/seat/mo) |
|---|---|---|---|---|
| Text chat (unlimited) | Yes | Yes | Yes | Yes |
| Ambient sound beds | 3 | All | All | All |
| Modes (Stillness/Heart/Focus/Flow/Rest/Sleep) | 1 (Stillness) | All 6 | All 6 | All 6 |
| /safety command | Always | Always | Always | Always |
| Basic reactions | Yes | Yes | Yes | Yes |
| Memory persistence across sessions | No | Yes | Yes | Yes |
| Constellation visualization | No | Yes | Yes | Yes |
| Growing tree visualization | No | Yes | Yes | Yes |
| Viral-share reflection images | No | Yes | Yes | Yes |
| Priority CloudFront routing | No | Yes | Yes | Yes |
| Voice mode (Sonnet 4.6) | No | No | Yes | Yes |
| Monthly licensed-adjacent counselor consult (15 min) | No | No | Yes | No |
| Extended memory (12 months) | No | No | Yes | Yes |
| Priority P50 <500ms routing | No | No | Yes | Yes |
| HIPAA BAA | No | No | No | Yes |
| Team analytics dashboard | No | No | No | Yes |
| Custom SSO | No | No | No | Yes |
| 24-hour support SLA | No | No | No | Yes |
The free tier carries a permanent, contractual-level commitment: no usage caps, no "you've used 3 of your 5 free conversations this month" banners, no paywall modals mid-session, no email drip sequences designed to manufacture FOMO. This commitment is documented in the public product charter and cannot be reversed without a public 90-day notice period.
The wellness app market has established a pricing band. Calm is $14.99/month or $69.99/year. Headspace is $12.99/month. BetterHelp, at the clinical end, runs $260–$360/month. Noom, further afield, runs $70/month.
Our Plus tier at $9/month is priced deliberately below Calm. This is a positioning signal, not a margin optimization. We want the first-time user evaluating the Plus tier to think "less than Calm" — signaling that we are a peer in the wellness space, not a premium product targeting a different segment. We are not underselling on quality; we are making an accessibility argument.
Pro at $19/month is priced at 2.1x Plus. This ratio is intentional and connects to the pricing psychology discussion in Section 8. The Pro tier exists in part as a decoy that makes Plus feel like the obvious rational choice for the majority of individual users.
Org at $99/seat with a 5-seat minimum ($495/month minimum commitment) is priced for institutional buyers who have procurement processes, compliance requirements, and dedicated wellness budgets. The comparison set here is not Calm — it is enterprise EAP programs ($30–$80/employee/month) and clinical platforms (Lyra Health, Spring Health, $200–$400/employee/month). At $99/seat, we are at the accessible end of the enterprise wellness market while being the only entrant with a HIPAA-compliant conversational AI layer.
---
These are assumptions, not observations. At current user count (pre-launch), we have no statistical basis for these numbers. They are sourced from analogous products (Replika, Woebot session data in published papers) and adjusted downward for conservatism.
| Cost Component | Free | Plus | Pro | Org (per seat) |
|---|---|---|---|---|
| LLM cost | $0.30 | $3.00 | $6.00 | $18.00 |
| Voice (Sonnet 4.6) | — | — | $2.00 | — |
| Infrastructure | $0.10 | $0.50 | $1.00 | $1.00 |
| Payment processing | — | $0.80 | $0.80 | $2.00 |
| Counselor referral | — | — | $5.00 | — |
| HIPAA ops overhead | — | — | — | $0.30 |
| Total COGS | $0.40 | $4.30 | $14.80 | $21.30 |
| Revenue | $0 | $9.00 | $19.00 | $99.00 |
| Gross Margin | (100%) | 52% | 22% | 78% |
The Free tier is a deliberate subsidy. At $0.40/user/month, a cohort of 10,000 free users costs $4,000/month — approximately $48,000/year. This is a marketing line item, not an operations failure. Every free user is a potential Plus conversion and a potential word-of-mouth referral. The free tier must be genuinely good, which means we must budget for it as a real cost.
Pro margin at 22% is the tightest. The $5 counselor referral cost is the primary driver. This cost will decrease as we achieve scale and renegotiate referral economics, and as we explore whether some Pro users opt out of the monthly consult (in which case the cost is zero for that user). The voice cost will also decrease as Anthropic's API pricing evolves. Pro at 22% gross margin is not unsustainable — it is the tier most likely to see margin expansion over time.
Org margin at 78% reflects institutional pricing power and the low marginal cost of adding enterprise users to an existing infrastructure base. Enterprise revenue will disproportionately fund free-tier subsidy and product development.
These conversion assumptions are targets derived from analogous SaaS products (Notion, Linear, Superhuman at early stage). They will be revised as real cohort data accumulates.
---
Retention modeling follows a modified Gompertz curve (standard for subscription SaaS). The critical assumption is that retention is higher for mental health-adjacent products than for productivity tools because the use case is emotional, not utilitarian — users who find genuine value form strong attachment. This is an optimistic assumption and we name it as such.
| Tier | Avg Retention Period | LTV Calculation | Estimated LTV |
|---|---|---|---|
| Free (if ever converts) | Blended | $9 × ~5.3 months avg paid tenure | $48 |
| Plus | 18 months | $9 × 18 months × 48% margin × retention curve (0.85 factor) | $180 |
| Pro | 24 months | $19 × 24 months × 22% margin × retention curve (0.90 factor) | $312 |
| Org (per account) | 36 months | $99 × 12 avg seats × 36 months × 78% margin × retention (0.88 factor) | $21,600 |
The retention factors (0.85, 0.90, 0.88) are applied to account for churn within the modeled period — they represent the probability the user is still subscribed at the average tenure month, not at the end. These are not observed churn rates. They are working assumptions derived from Baremetrics industry benchmarks for consumer subscription apps (B2C SaaS median monthly churn: 4–7%) and enterprise SaaS (median monthly churn: 0.5–2%).
| Segment | CAC Target | Payback Period | Rationale |
|---|---|---|---|
| Plus/Pro | <$60 | 12 months (Plus), 8 months (Pro) | Performance marketing, content, word-of-mouth |
| Org | <$2,500 | 3 months | Direct sales, partnership channel, EAP referrals |
A Plus LTV of $180 at a CAC of $60 yields an LTV:CAC ratio of 3:1, the standard SaaS health benchmark. A Pro LTV of $312 at $60 CAC yields 5.2:1 — healthier, but only meaningful once we validate the 24-month retention assumption. Org at $21,600 LTV and $2,500 CAC yields 8.6:1 — strong, and typical for enterprise deals with low churn.
At current user count (pre-launch), none of these figures have statistical grounding. We have no retention data. We have no conversion funnel data. We have no churn data. These are modeling inputs, not observed metrics. The purpose of stating them now is to have a clear target set against which actual cohort data can be measured. Investors and internal stakeholders should treat all LTV and CAC figures in this memo as aspirational benchmarks with wide confidence intervals until we have 6+ months of real cohort data.
---
Stripe handles all web and direct subscriptions. Standard rate: 2.9% + $0.30 per transaction. Under Stripe Tax, VAT and sales tax collection is automated for 35+ countries — this is load-bearing for any European or Canadian user base.
Stripe Billing manages subscription lifecycle: creation, upgrades, downgrades, pro-rata adjustments, dunning (failed payment recovery), and cancellation. We configure soft dunning (3 retry attempts over 7 days before cancellation) rather than aggressive dunning, consistent with our ethical stance.
Stripe Radar provides fraud detection. We accept the default risk threshold — we do not tighten it to squeeze margins, because false positives on legitimate users are a product quality issue, not just a financial one.
Apple IAP is required for any in-app subscription on iOS — Apple enforces this as a platform policy. Under the Apple Small Business Program (revenue under $1M), the commission rate is 15% rather than the standard 30%. We qualify at launch and must maintain this status. Year 2 revenue projections do not exceed the $1M threshold; Year 3 may, at which point this assumption must be revisited and pricing adjusted.
Google Play Billing applies the same 15% commission for the first $1M in annual revenue under the Google Play reduced service fee program.
The practical implication: mobile-originated subscriptions have lower gross margin than web-originated subscriptions by approximately 12–15 percentage points. We account for this in blended margin projections by assuming a 40/60 mobile/web subscription split (assumption; to be validated).
For Org-tier enterprise customers, particularly those in jurisdictions with complex tax treatment or foreign exchange requirements, Paddle or LemonSqueezy serve as merchant-of-record. This shifts tax compliance burden for international sales and simplifies enterprise procurement in markets like Germany, France, Japan, and Australia. The fee is approximately 5% of transaction value.
Annual subscriptions are offered at 2 months free equivalent: Plus $90/year (vs. $108 annualized monthly), Pro $190/year (vs. $228 annualized monthly). This discount structure serves cash flow (upfront revenue), reduces churn (users who paid annually have lower monthly cancel probability), and is a clean, non-deceptive offer — the math is transparent.
Pro-rata refund within 30 days of any subscription charge, no questions asked. No refund after 30 days for the current billing period. This policy is published in plain language on the pricing page and in the onboarding flow. We do not bury it.
---
Each rejected model is documented with reasoning, not dismissed. Future teams should be able to read this section and understand why we chose not to pursue options that appear financially attractive.
Advertising. The mirror cannot have advertisers. Full stop. There is no version of Silent Infinity where a user's emotional state, conversation topics, or behavioral signals are sold to third parties or used to target them with ads. This applies to retargeting (using Silent Infinity behavior to target users on other platforms) as well as in-product advertising. Revenue: $0. Policy: permanent.
Usage-based pricing ($0.25/conversation). On paper this looks like a "pay for what you use" fairness argument. In practice it creates anxiety around the product — the last thing a user in distress needs is to be calculating whether they can afford to finish this session. Usage-based pricing also creates an adversarial dynamic in the exact moments when the product is most valuable (high-frequency use during acute distress). This model incentivizes exactly what we are trying to prevent.
Freemium with dark patterns. We have observed the Duolingo streak mechanic, the Calm "your streak will expire" notification, the Headspace "you're so close to your monthly goal" modal, and the Noom guilt-inducing check-in prompt. We refuse all of them. The distinction between engagement features and manipulation features is real and we will enforce it. A feature that creates genuine value for the user is fine. A feature that creates anxiety or guilt to drive re-engagement is not fine. This is not a fuzzy line in most cases.
Data sale or anonymized research licensing. We cannot sell user data, full stop — this is foreclosed by our privacy commitment and by the nature of the conversations. Even "anonymized" conversation data from a mental health-adjacent product raises serious re-identification risks and ethical questions about consent that we are not willing to navigate. Revenue: $0. Policy: permanent.
Token sales or crypto integration. Brand misalignment. Silent Infinity is a product of stillness and trust. Associating it with speculative financial instruments, even tangentially, undermines the brand and the user relationship. We have no interest in this.
Sponsored content (wellness supplement partnerships, etc.). Identical problem to advertising. If a supplement brand sponsors a "sleep hygiene" conversation flow, the conversation is no longer neutral. The user trusts us to be disinterested. The moment we take money to have an opinion, we have broken that trust.
Employer wellness kickbacks (B2B2C without disclosure). We are open to B2B2C arrangements where employers subsidize or pay for employee access to Silent Infinity. This is legitimate and we describe the mechanism in Section 7. However: any B2B2C arrangement must be fully transparent to the end user. If an employer is paying for the subscription, the user sees that. We do not launder corporate payments into an apparently neutral product experience.
Clinical referral commissions (undisclosed). If we create pathways from Silent Infinity conversations to licensed therapists or psychiatrists, there is a business model where we take a referral fee from those providers. This is allowed only under explicit conditions: the user must be informed of the referral relationship in plain language before any referral is made, the user must consent, and the referral must be to providers we have independently vetted. Undisclosed referral commissions are not acceptable.
---
Subscription — primary revenue stream, target 80% of total revenue. This is the core and it must remain the core. If subscription revenue falls below 70% of total, we should treat that as a flag to rebalance.
Enterprise seat licensing (Org tier) — target 15%. Institutional buyers, EAP programs, university wellness centers, corporate HR platforms. Revenue here should scale with user count, not with engagement depth.
Donation tier (Innovation 6 pathway) — target 2%. The crisis layer of Silent Infinity — the /safety command and its underlying capabilities — should ultimately be funded in a way that is structurally independent from commercial revenue. A 501(c)(3) vehicle for this layer enables grants, donations, and philanthropic funding. Revenue generated through this channel is reinvested in the crisis layer, not into the commercial product. Target: 2% of total funding comes from philanthropic sources by Year 3.
Wisdom-library partnerships — target 3%. Revenue share arrangements with estates and publishers for authorized translations, audiobook partnerships, or integration of curated wisdom texts (Marcus Aurelius, Thich Nhat Hanh, Epictetus). These are content partnerships, not sponsorships — we pay for the content license or share revenue on derivative products. The content must be selected on editorial merit, not on which estate offers the best terms.
Research grants — 0% profit by policy. If Silent Infinity becomes a research partner for a university study on digital mental health interventions, any grant revenue goes entirely to covering the cost of the research participation (data infrastructure, consent management, researcher access). No profit extraction from research partnerships. This is both an ethical commitment and a legal/IRB requirement in any formal research context.
---
The three-tier structure (Plus $9, Pro $19, Org $99) is designed with explicit awareness of the decoy effect (Ariely, Predictably Irrational, 2008; Huber, Payne & Puto, 1982). Pro at $19 exists partly as a decoy that makes Plus at $9 feel like the obviously rational choice for individual users. The majority of individual users will look at the tier list and conclude: "I don't need the counselor consult or voice mode, $9 is the right answer." That is the intended outcome.
This is not manipulative because all tiers deliver genuine value at their price point. The decoy effect only becomes ethically problematic when the decoy tier is a hollow option designed purely to shift choices — we reject that version. Pro must be a real, valuable product for users who need voice mode or the counselor touchpoint.
Hermann Simon in Confessions of the Pricing Man (2015) is categorical: price is a signal before it is a cost. Our Plus tier at $9/month communicates "we are a peer of Calm and Headspace, not a budget knockoff, but we are accessible." If we priced Plus at $4.99, we would signal lower quality. At $9, we are $6 less than Calm while implying equivalent or superior quality. This is intentional anchoring, not accidental pricing.
We do not run "Black Friday 50% off" promotions. We do not show countdown timers. We do not fabricate original prices ("was $19, now $9!") because the "original price" was never real. These tactics work in the short term and erode brand trust over the medium term. Silent Infinity users are, by design, people who have chosen a product that does not manipulate them. Discount theater is a betrayal of that implicit contract.
Annual prepay at 2 months free is an honest discount — the math is transparent, the savings are real, and there is no artificial urgency.
Subscribers on annual plans keep their tier price at the rate they locked in, permanently. If we raise prices in Year 2, existing annual subscribers are never affected. New subscribers see the new price. This is a brand commitment, not a marketing gimmick — it signals that we treat long-term users as partners, not as captive revenue.
$3/month, honor system, no verification required, no embarrassing documentation process. Anyone who needs the reduced price can select it. We will lose some revenue to users who could afford more. We accept this because: (a) the marginal revenue loss is small at our scale, (b) accessibility is a core value, and (c) a verification process would be degrading and inconsistent with how we treat users. We trust people.
---
| Metric | Year 1 | Year 2 | Year 3 |
|---|---|---|---|
| Total DAU | 1,000 | 25,000 | 150,000 |
| Paid conversion rate | 4% | 5% | 6% |
| Paid individual users | 40 | 1,250 | 9,000 |
| Org accounts | 0 | 3 | 20 |
| Avg seats/Org account | — | 10 | 12 |
| Blended individual ARPU | $10.50 | $10.50 | $11.00 |
| Individual MRR | $420 | $13,125 | $99,000 |
| Org MRR | $0 | $2,970 | $23,760 |
| Total MRR | ~$420 | ~$16,095 | ~$122,760 |
| ARR | ~$5,000 | ~$193,000 | ~$1.47M |
Note: The original brief projected $3,600 MRR in Year 1 (1,000 DAU, 4% conversion). Our model produces $420 MRR at the same DAU and conversion rate because we used blended ARPU of $10.50 applied to 40 paid users — the original brief appears to have used $90/user, which may reflect annual pricing assumptions. We use monthly ARPU throughout for consistency; ARR is MRR × 12.
Year 3 ARR of ~$1.47M is below the original $9.7M target. The gap is attributable to Org account count and seat assumptions — our 20 accounts × 12 seats × $99 generates $285,000 ARR, whereas the brief's 20 Org accounts at implied higher volumes could generate more. These projections should be revisited when we have enterprise sales pipeline data.
At current burn rate assumptions (3-person team, infrastructure, legal, minimal marketing), monthly operating costs are approximately $25,000–$40,000. Break-even at $35,000 MRR target requires approximately 3,300 paid individual users at blended $10.50 ARPU, or mix of Org accounts. On the current trajectory, this is achievable in month 20–24 — consistent with the "month 18 at current pace" estimate in the brief, with the caveat that Org deal timing is the dominant variable.
A 1% increase in conversion rate (from 6% to 7% in Year 3) on 150,000 DAU adds 1,500 paid users at $10.50 ARPU = $15,750/month additional MRR = $189,000 additional ARR. The brief's $2.4M figure is achievable only if blended ARPU is significantly higher (suggesting a higher Pro mix) or Org accounts are substantially more numerous. We flag this as an area for model reconciliation once Year 1 data arrives.
---
Pricing changes are one of the highest-trust moments in a consumer product relationship. Users notice. The following guardrails are not aspirational — they are operational constraints.
Grandfathered users: 12-month minimum notice before any price change. If we decide to raise the Plus price from $9 to $11, users on annual lock receive 12 months of notice before any change affects their renewal. Monthly subscribers receive 60 days. No exceptions.
No "pay to skip" mechanics. We will never charge users to remove a dark pattern, because we will never have dark patterns to remove. A product that first installs irritants and then sells relief from those irritants is extracting value it never created. We will not build that product.
No dynamic or personalized pricing based on wealth signals. We will not charge users more based on their location, inferred income, device type, or behavioral signals. A user on a new iPhone 16 Pro Max pays the same as a user on a five-year-old Android. Personalized pricing, while legal in most jurisdictions, is inconsistent with our values model.
Annual transparency. The quarterly transparency report will include: current tier pricing, cost-of-service data at the blended level, gross margin by tier, and a plain-language explanation of any pricing changes made in the period and why. Users deserve to understand the economics of the product they are paying for.
---
Every pricing change — including tier additions, removals, price adjustments, and feature moves between tiers — must be documented using the following template before implementation.
---
Decision ID: [Sequential, e.g., PRICE-2026-001]
Date Proposed: [YYYY-MM-DD]
Proposed Change: [One paragraph, plain language description of the specific change]
Rationale: [What problem does this solve? What data or principle drives this decision?]
Projected Financial Impact: [MRR delta, ARR delta, affected user count, margin delta]
Customer Impact Analysis: [Who is affected? How? Are any users worse off? If yes, how are we mitigating?]
Alternatives Considered: [At least two alternatives with brief rationale for rejection]
30-Day Comment Period: [Open date — close date. Summary of feedback received]
Final Decision: [Proceed / Modify / Reject, with one-paragraph rationale]
Implementation Date: [YYYY-MM-DD]
Responsible Owner: [Name / role]
Post-Implementation Review Date: [90 days after implementation]
---
This template exists because pricing decisions affect real people's finances and their trust in the product. A decision log is both a governance document and a forcing function — the act of writing a customer impact analysis before implementing a change catches problems that post-hoc rationalization misses.
---
1. Slywotzky, A. J., Morrison, D. J., Moser, B., Mundt, K., & Quella, J. (2003). Profit Patterns: 30 Ways to Anticipate and Profit from Strategic Forces Reshaping Your Business. Times Business. — Core framework for revenue model selection (Ad Model vs. Engagement Model vs. Customer Relationship Model).
2. Simon, H. (2015). Confessions of the Pricing Man: How Price Affects Everything. Springer. — Foundational reference on price as signal, anchoring strategy, and the psychology of tier design.
3. Ramanujam, M., & Tacke, G. (2016). Monetizing Innovation: How Smart Companies Design the Product Around the Price. Wiley. — Framework for pricing before building; validates the decision to fix tier prices before product launch rather than after.
4. Heath, C., & Heath, D. (2013). Decisive: How to Make Better Choices in Life and Work. Crown Business. — Widening the option set; used here in the context of evaluating rejected revenue models rigorously rather than defaulting to the first viable structure.
5. Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux. — Anchoring and adjustment (Chapter 11); theoretical foundation for competitive anchoring against Calm and Headspace as reference prices.
6. Huber, J., Payne, J. W., & Puto, C. (1982). "Adding Asymmetrically Dominated Alternatives: Violations of Regularity and the Similarity Hypothesis." Journal of Consumer Research, 9(1), 90–98. — Original empirical study on the attraction/decoy effect; informs the Plus/Pro tier structure.
7. Baremetrics (2025). SaaS Benchmarks: Churn, LTV, and CAC by Company Size. — Industry benchmarks used to calibrate retention curve assumptions. Note: figures used are 2024–2025 cohort data; should be refreshed annually.
---
End of Memorandum
Document Classification: Internal Working Document
Next Review Date: 2026-07-01 (prior to any Series A materials)
Maintained by: TITAN / Harnoor Singh