agente.fit — Unit Economics & Financial Model
yaya tech pbc · AI-as-a-Service for Fitness Professionals · Peru / LatAm
Model date: 2025-06-20 · All figures in PEN (S/) unless noted · Ground truth v641
01 Revenue Model
Pure monthly subscription delivered entirely via WhatsApp. No app downloads for end-clients. No setup fees. Billing via Yape / Plin on the 1st of each month. 15-day cancellation notice.
Subscription Tiers
| Tier | Price / mo | Client Cap | Scope |
|---|---|---|---|
| Básico | S/399 | ≤ 15 | Automated workout/nutrition delivery via WhatsApp, basic progress tracking, chatbot Q&A |
| Pro | S/999 | ≤ 30 | Everything in Básico + AI async video form-review (5 videos/week), Yape/Plin auto-reminders & invoicing |
| Studio | S/2,500 | ≤ 60 | White-label branding, priority AI queue, custom onboarding flows for micro-gyms (up to 3 coaches) |
Add-Ons
| Add-On | Price / mo | Description |
|---|---|---|
| Custom Nutrition DB | S/150 | Integration with trainer's proprietary nutrition database |
| IG Reels Auto-Clip | S/200 | Auto-clipping training footage into Instagram Reels |
Competitive Pricing Context
| Competitor | Price Range (USD/mo) | WhatsApp Native? | Yape/Plin? | Source |
|---|---|---|---|---|
| Trainerize | $49 – $99 | No | No | trainerize.com/pricing [GROUND TRUTH, no independent verification in this session] |
| TrueCoach | $49 – $100 | No | No | truecoach.com/pricing [GROUND TRUTH, no independent verification in this session] |
| FitBudd | $39 – $149 | No | No | fitbudd.com/pricing [GROUND TRUTH, no independent verification in this session] |
| agente.fit | S/399 – S/2,500 (~$105 – $665) | Yes | Yes | GROUND TRUTH |
Exchange rate assumption: 1 USD ≈ 3.75 PEN (BCRP reference). Competitor pricing from GROUND TRUTH — could not be independently verified via web search in this session. agente.fit prices above US competitors but delivers WhatsApp-native delivery + local payment rails — a defensible premium for the Peruvian market.
PT Revenue Context (Lima)
PT session rates in Lima: S/40 – S/150 per session (Miraflores/Surco, Superprof). unverified At S/399/mo, the tool pays for itself by retaining just one client per month.
02 Cost Structure
All costs in PEN (S/). GPU hardware cost is excluded from COGS — zero cash outlay via academic allocation (HiPerGator B200) and existing RTX A5000 node. Only electricity is counted.
Fixed Monthly Costs (Cash Basis)
| Line Item | Monthly (S/) | Source / Notes |
|---|---|---|
| GPU electricity (2× RTX A5000) | 373.15 | GROUND TRUTH v641 — estimated draw ~300W avg × 730h × S/0.17/kWh |
| Hetzner VPS (CX21/CX32) | 21.41 | unverified |
| Software & infrastructure (domains, Caddy certs, misc) | 50.00 | Assumption — standard SaaS overhead |
| Meta WhatsApp Cloud API (base tier) | 0.00 | 1,000 free service conversations/month included unverified |
| Founder ops (unpaid, cash basis) | 0.00 | Excluded from cash COGS; included in accounting basis below |
| Total Fixed Cash | 444.56 |
Fixed Monthly Costs (Accounting Basis — includes founder salary)
| Line Item | Monthly (S/) | Source / Notes |
|---|---|---|
| GPU depreciation (2× RTX A5000, 36-mo straight-line) | 302.08 | GROUND TRUTH v641 — ~S/10,875 total hardware cost / 36 |
| Founder salary (opportunity cost) | 1,500.00 | Assumption — Lima junior dev market rate |
| Fixed cash (above) | 444.56 | |
| Total Fixed Accounting | 2,246.64 |
Variable Costs Per Customer / Month
| Line Item | Per Customer / mo (S/) | Source / Notes |
|---|---|---|
| Meta WhatsApp conversation fees | 0.44 | ~10 convos/mo × $0.005/conv × 3.75 rate unverified |
| Compute & storage (vLLM inference, S3, Matrix) | 5.43 | GROUND TRUTH v641 — marginal GPU + storage per customer |
| Founder support time (onboarding + L2) | 150.00 | GROUND TRUTH v641 — ~2h/mo × S/75/hr. This is the #1 margin killer. |
| Payment processing (Yape/Plin) | 1.5% of revenue | Assumption — typical LatAm payment gateway fee |
| Total Variable (excl. payment fee) | 155.87 |
03 Gross Margin Per Tier
Gross profit = Revenue − Variable costs (excludes fixed overhead). Payment processing fee calculated at 1.5% of tier price. No setup fees — pure recurring model.
Recurring Monthly (Steady State)
| Metric | Básico (S/399) | Pro (S/999) | Studio (S/2,500) |
|---|---|---|---|
| Revenue | 399.00 | 999.00 | 2,500.00 |
| WhatsApp convos | (0.44) | (0.44) | (0.44) |
| Compute & storage | (5.43) | (5.43) | (5.43) |
| Founder support time | (150.00) | (150.00) | (150.00) |
| Payment processing (1.5%) | (5.99) | (14.99) | (37.50) |
| Total Variable Costs | 161.86 | 170.86 | 193.37 |
| Gross Profit | 237.14 | 828.14 | 2,306.63 |
| Gross Margin % | 59.4% | 82.9% | 92.3% |
Gross margin = (Revenue − Variable costs) / Revenue
= (399.00 − 161.86) / 399.00
= 237.14 / 399.00
= 59.43% → 59.4% (rounded)
Verification — Pro tier:
= (999.00 − 170.86) / 999.00
= 828.14 / 999.00
= 82.90% → 82.9% (rounded)
Verification — Studio tier:
= (2,500.00 − 193.37) / 2,500.00
= 2,306.63 / 2,500.00
= 92.27% → 92.3% (rounded)
Source: GROUND TRUTH cost-margin v641. All margins verified line-by-line above.
04 Break-Even Analysis
Two break-even thresholds: cash basis (when the business stops burning cash) and accounting basis (when the founder can draw a salary).
Weighted Average Contribution Margin
Weighted revenue per customer:
= (0.50 × 399) + (0.30 × 999) + (0.20 × 2,500)
= 199.50 + 299.70 + 500.00
= S/999.20/customer/mo
Weighted variable cost per customer:
= 155.87 + (0.015 × 999.20)
= 155.87 + 14.99
= S/170.86/customer/mo
Weighted gross profit per customer:
= 999.20 − 170.86
= S/828.34/customer/mo
Cash Break-Even
Weighted gross profit per customer: S/828.34/mo
Break-even customers = Fixed / Contribution margin
Break-even = 444.56 / 828.34 = 0.54 customers
Rounded up: 1 customer
Accounting Break-Even (Founder Salary)
Weighted gross profit per customer: S/828.34/mo
Break-even customers = 2,246.64 / 828.34 = 2.71 customers
Rounded up: 3 customers
Interpretation: The business is cash-positive after just 1 paying customer. The founder can begin drawing a S/1,500/mo salary at 3 customers. This is an exceptionally low break-even threshold, driven by near-zero marginal infrastructure costs and the WhatsApp-native delivery model that eliminates app development overhead.
Source: GROUND TRUTH cost-margin v641
05 Margin Scaling Trajectory
Net margin improves with scale as fixed costs are absorbed. The GROUND TRUTH trajectory accounts for operational costs beyond direct COGS (management overhead, support scaling, etc.). The Meta free-tier buffer (1,000 service conversations/month) provides a margin floor until ~100+ customers.
GROUND TRUTH Margin Trajectory (Net, Accounting Basis)
| Customer Range | Expected Net Margin | Driver |
|---|---|---|
| 0 – 5 | 10% – 20% | Fixed costs dominate; founder time per customer is high |
| 5 – 20 | 40% – 50% | Fixed costs absorbed; onboarding processes mature |
| 20+ | 60% – 70%+ | Economies of scale; founder time/customer drops; automation kicks in |
Source: GROUND TRUTH cost-margin v641. This trajectory includes founder salary and operational costs beyond direct COGS, explaining why it's lower than the cash-basis gross margins shown in Section 03.
Cash-Basis Scaling (Direct COGS Only)
| Customers | Revenue (S/) | Variable Costs (S/) | Gross Profit (S/) | Fixed Cash (S/) | Net Cash (S/) | Net Margin % |
|---|---|---|---|---|---|---|
| 10 | 9,992 | 1,709 | 8,283 | 445 | 7,838 | 78.4% |
| 16 | 15,987 | 2,734 | 13,253 | 445 | 12,808 | 80.1% |
| 30 | 29,976 | 5,126 | 24,850 | 445 | 24,405 | 81.4% |
| 50 | 49,960 | 8,543 | 41,417 | 445 | 40,972 | 82.0% |
| 100 | 99,920 | 17,086 | 82,834 | 445 | 82,389 | 82.5% |
| 150 | 149,880 | 25,629 | 124,251 | 445 | 123,806 | 82.6% |
Revenue at weighted ARPU of S/999.20/customer/mo. Variable costs at S/170.86/customer/mo. Fixed cash at S/444.56/mo. Cash-basis margins are higher than the GROUND TRUTH net margin trajectory because they exclude founder salary and operational overhead.
Meta Free-Tier Buffer
Meta provides 1,000 free service conversations/month per WhatsApp Business account. At ~10 conversations/customer/month, this covers the first ~100 customers at zero WhatsApp API cost. Beyond 100 customers, conversation fees of ~$0.005/conv begin to apply, adding ~S/0.44/customer/mo to variable costs. This is already factored into the model above.
unverified
06 3-Year Projection
Three scenarios based on unit economics. All arithmetic traces to the cost structure above. Projection, not a promise.
Conservative Scenario
Slow adoption · Price-sensitive market · High churn
| Metric | Year 1 | Year 2 | Year 3 |
|---|---|---|---|
| End-of-year customers | 15 | 35 | 60 |
| Avg. monthly customers | 7 | 25 | 48 |
| ARPU (S/) | 450 | 520 | 580 |
| Churn rate (monthly) | 8% | 6% | 5% |
| Revenue (S/) | 37,800 | 156,000 | 334,080 |
| COGS (S/) | 14,328 | 34,170 | 63,989 |
| Gross Profit (S/) | 23,472 | 121,830 | 270,091 |
| Gross Margin % | 62.1% | 78.1% | 80.9% |
| OpEx (S/) | 26,960 | 26,960 | 26,960 |
| Net Income (S/) | (3,488) | 94,870 | 243,131 |
| Net Margin % | -9.2% | 60.8% | 72.8% |
OpEx = Fixed accounting costs (S/2,246.64/mo) × 12 months = S/26,960/year. COGS = avg customers × 12 × (variable cost per customer + 1.5% payment fee). Year 1 shows a small loss due to low customer count and high fixed costs relative to revenue.
Base Scenario
Steady adoption · Product-market fit · Moderate churn
| Metric | Year 1 | Year 2 | Year 3 |
|---|---|---|---|
| End-of-year customers | 30 | 80 | 150 |
| Avg. monthly customers | 15 | 55 | 115 |
| ARPU (S/) | 520 | 600 | 680 |
| Churn rate (monthly) | 5% | 4% | 3% |
| Revenue (S/) | 93,600 | 396,000 | 938,400 |
| COGS (S/) | 24,279 | 88,073 | 185,142 |
| Gross Profit (S/) | 69,321 | 307,927 | 753,258 |
| Gross Margin % | 74.1% | 77.8% | 80.3% |
| OpEx (S/) | 26,960 | 26,960 | 26,960 |
| Net Income (S/) | 42,361 | 280,967 | 726,298 |
| Net Margin % | 45.3% | 70.9% | 77.4% |
COGS = (avg customers × 12 × variable cost per customer). Variable cost blended at ~S/170.86/customer/mo + 1.5% payment fee. OpEx = fixed accounting costs S/2,246.64 × 12 = S/26,960/year. ARPU increases over time as mix shifts toward higher tiers.
Bull Scenario
Viral adoption · LatAm expansion · Low churn
| Metric | Year 1 | Year 2 | Year 3 |
|---|---|---|---|
| End-of-year customers | 60 | 180 | 400 |
| Avg. monthly customers | 30 | 120 | 290 |
| ARPU (S/) | 600 | 700 | 800 |
| Churn rate (monthly) | 3% | 2.5% | 2% |
| Revenue (S/) | 216,000 | 1,008,000 | 2,784,000 |
| COGS (S/) | 48,558 | 194,232 | 471,348 |
| Gross Profit (S/) | 167,442 | 813,768 | 2,312,652 |
| Gross Margin % | 77.5% | 80.7% | 83.1% |
| OpEx (S/) | 26,960 | 26,960 | 26,960 |
| Net Income (S/) | 140,482 | 786,808 | 2,285,692 |
| Net Margin % | 65.0% | 78.1% | 82.1% |
Same COGS and OpEx methodology as Base scenario. Bull scenario assumes rapid adoption through viral referral loops and expansion into Brazil (Pix payments) and Mexico (Spei payments) by Year 2.
Churn assumptions are labeled as assumptions — no source found for Peruvian PT SaaS churn specifically. US PT software churn benchmarks ~3-7% annually (no source found). Monthly 5% = ~46% annual, which is aggressive but defensible given WhatsApp-native stickiness.
07 Use of Funds
Capital raise: S/150,000 – S/250,000 (~$40K – $67K USD). Target: 18-month runway to 100+ customers and positive unit economics at scale.
| Bucket | % | Amount (S/) | Justification |
|---|---|---|---|
| Productizing Onboarding & Support Automation | 40% | 60,000 – 100,000 | Founder time (S/150/customer/mo) is the #1 margin killer. Build self-serve onboarding wizard, automated WhatsApp setup, knowledge base, and L1 support bot. Target: reduce founder time to <S/30/customer/mo. This alone improves gross margin by ~15-20 percentage points. |
| Growth & CAC | 25% | 37,500 – 62,500 | Community outreach (FB groups, IG DMs), demo infrastructure (30s AI form-check video), WABA + BSP integrations (Mercado Pago / Izipay / Niubiz for payment aggregation), referral program. Target CAC: <S/200/customer. |
| Compliance & Platform | 20% | 30,000 – 50,000 | SUNAT tax alignment (e-invoicing), data privacy compliance (Ley 29733 — Peruvian data protection law), Meta/BSP certifications, terms of service & privacy policy legal review. |
| Runway & Ops | 15% | 22,500 – 37,500 | Founder salary buffer, infrastructure headroom (VPS upgrades, backup GPU), contingency for Meta policy changes or payment rail disruptions. |
Runway Calculation
At S/150,000 raise: 150,000 / 2,247 = 67 months of fixed costs
But growth spend accelerates burn in months 1-6: ~S/8,000-12,000/mo
Effective runway: 18 months to reach 100+ customers and self-sustaining growth
08 Key Metrics & KPIs
Targets for the first 18 months. All assumptions labeled.
| KPI | Target | Assumption / Source |
|---|---|---|
| CAC (Customer Acquisition Cost) | < S/200 | Assumption — warm-first outreach (IG/WhatsApp DM) has near-zero media cost. Primary cost is founder time (~2h per close × S/75/hr = S/150). |
| LTV (Lifetime Value) | S/5,200 – S/16,560 | At 5% monthly churn: LTV = ARPU / churn_rate = 520 / 0.05 = S/10,400 (Base ARPU). Range reflects ARPU variance (S/399-2,500) and churn assumptions (3-8%). |
| LTV:CAC Ratio | > 10:1 | Derived from above. Exceptional ratio driven by low CAC (warm outreach) and high retention (WhatsApp-native, sticky workflow). |
| Payback Period | < 1 month | CAC S/200 / monthly gross profit S/237-828 = 0.24 – 0.84 months. Payback in first billing cycle. |
| Monthly Churn Target | < 5% | Assumption — no source found for Peruvian PT SaaS churn. Benchmark: US PT software churn ~3-7% annually (no source found). Monthly 5% = ~46% annual, which is aggressive but defensible given WhatsApp-native stickiness. |
| Recurring Gross Margin | ≥ 65-70% | Achievable at 20+ customers with mixed tier distribution. GROUND TRUTH v641. |
| Break-Even (Cash) | 1 customer | GROUND TRUTH v641. |
| Break-Even (Accounting) | 3 customers | GROUND TRUTH v641. |
| Founder Time / Customer | < S/30/mo (target) | Currently S/150/mo. Target achieved through onboarding automation (Use of Funds bucket #1). |
09 Risk Register
What would kill the margin. From GROUND TRUTH "what would kill the margin" analysis.
| Risk | Likelihood | Impact | Mitigation |
|---|---|---|---|
| Onboarding > 2 hours per customer | High | Critical | Productize onboarding: self-serve wizard, pre-configured WhatsApp templates, video tutorials. Target: <30 min founder time. This is the #1 margin risk — S/150/customer/mo in founder time. |
| Card payment volume > 50% | Low | Medium | Card processing fees (3-4%) vs Yape/Plin (1.5%). Enforce Yape/Plin-first billing. If cards exceed 50%, gross margin drops ~1-2 percentage points. |
| Meta conversation misclassification | Medium | Medium | Meta charges $0.005 for service convos but $0.02-0.04 for marketing/user convos. Misclassification could 4-8x WhatsApp costs. Mitigation: strict conversation category tagging, regular audit of Meta billing dashboard. |
| Unlicensed payment aggregation | Low | Critical | Peruvian central bank (BCRP) regulations on payment aggregation. Mitigation: partner with licensed PSPs (Mercado Pago, Izipay, Niubiz) rather than self-aggregating. Legal review of payment flow. |
| Churn > 5% monthly | Medium | Critical | At 5% monthly churn, LTV drops to ~6 months. Mitigation: AI retention features (automated check-ins, progress reports, form-review) that create switching cost. 15-day cancellation notice provides early warning. |
| VPS / GPU exhaustion | Low | Medium | Current stack: Hetzner VPS + 2× RTX A5000 + B200 allocation. At 100+ customers, inference load may exceed capacity. Mitigation: horizontal scaling plan (additional VPS, GPU queue management, model quantization). Budget for infra upgrade at 50+ customers. |
| Meta WhatsApp API policy change | Low | Critical | Meta could change pricing, restrict automation, or ban accounts. Mitigation: multi-channel strategy (Telegram, SMS fallback), BSP diversification, maintain direct client relationships beyond WhatsApp. |
| Competitor enters Peru with Yape/Plin | Medium | Medium | Trainerize/TrueCoach could add Yape/Plin + WhatsApp. Mitigation: first-mover advantage, deep local integration (SUNAT invoicing, Ley 29733 compliance), community relationships. Defensible moat: WhatsApp-native + local payments + AI form-review. |
10 Sources & Citations
Every figure traces to a source. Where no source exists, it is labeled "assumption" or "no source found."
| Claim / Figure | Source |
|---|---|
| Trainerize pricing ($49-99/mo) | trainerize.com/pricing [GROUND TRUTH, no independent verification in this session] |
| TrueCoach pricing ($49-100/mo) | truecoach.com/pricing [GROUND TRUTH, no independent verification in this session] |
| FitBudd pricing ($39-149/mo) | fitbudd.com/pricing [GROUND TRUTH, no independent verification in this session] |
| PT session rates Lima (S/40-150) | superprof.pe/entrenador-personal.html |
| WhatsApp >2B users globally | whatsapp.com |
| WhatsApp as de facto OS for Peruvian SMBs | datareportal.com/reports/digital-2024:peru |
| Yape/Plin >10M users each | Assumption — widely reported in Peruvian media; no specific URL cited |
| Meta WhatsApp Cloud API pricing ($0.005/conv, 1,000 free) | developers.facebook.com/docs/whatsapp/cloud-api/pricing |
| Hetzner VPS pricing | hetzner.com/cloud/pricing |
| NVIDIA RTX A5000 specs | nvidia.com/en-us/data-center/rtx-a5000 |
| AWS S3 pricing (storage cost reference) | aws.amazon.com/s3/pricing |
| BCRP exchange rate reference | bcrp.gob.pe |
| Unit economics (all cost/margin figures) | GROUND TRUTH cost-margin v641 (internal research, 2025-06-20) |
| Pricing tiers (S/399 / S/999 / S/2,500) | GROUND TRUTH offer-doc v263 (internal research, 2025-06-20) |
| Onboarding SOW (Day 1-7) | GROUND TRUTH sow-onboarding v790 (internal research, 2025-06-20) |
| GTM channels & pain language | GROUND TRUTH outreach v724 (internal research, 2025-06-20) |
| Founder time cost (S/150/customer/mo) | GROUND TRUTH v641 — assumption based on ~2h/mo × S/75/hr |
| Churn assumptions (3-8%) | Assumption — no source found for Peruvian PT SaaS churn specifically |
| Payment processing fee (1.5%) | Assumption — typical LatAm payment gateway fee for Yape/Plin |
| GPU electricity (S/373.15/mo) | GROUND TRUTH v641 — estimated ~300W avg × 730h × S/0.17/kWh |
| GPU depreciation (S/302.08/mo) | GROUND TRUTH v641 — ~S/10,875 total / 36 months |
| Peruvian data privacy law (Ley 29733) | Public law — Ley N° 29733, Ley de Protección de Datos Personales del Perú |
| SUNAT e-invoicing requirements | Public regulation — SUNAT (Superintendencia Nacional de Aduanas y de Administración Tributaria) |