The single source of truth for the LUMÉ network of aesthetic-cosmetology studios. Every operational project and all the automation take their numbers from it. The company is fictional — the economics are assembled from real Ukrainian data for 2026.
LUMÉ does not exist. A fictional network — like "Kontur" in movements I–VII. But the economics aren't "fitted to corridors": the market parameters (prices, rent, salaries, taxes, CPL) are taken from real sources and gathered in research/source_register.md with confidence grades.
Every number is marked by origin: S a real market input · M my model assumption · D derived from S and M.
The 2021–2026 retrospective is a simulation of what a network of this type would have lived through (the macro is real, grade A; the LUMÉ trajectory is M), not "LUMÉ's history."
The headline numbers (LTV, CAC, margin, utilisation) are computed, not measured — from real inputs via params.py → model.py → figures.json. Change any input and they recompute. The skill here isn't "state a number" but build the machine that derives it from checkable data, with every assumption on show.
A network of medical aesthetic-cosmetology studios in Kyiv. 4 operating studios (Pechersk, Podil, Obolon, Pozniaky), a fifth opening in the centre. An LLC, a VAT payer, an MoH licence. ~16 treatment rooms, ~20 chairside practitioners (3 cosmetologists + 2 doctors per studio) plus admins, nurses, managers.
The position is "the economics of trust" (the portfolio's through-line thesis): in a high-risk category the one who removes the fear first wins. The diagnostic consultation is paid and not credited toward the procedure — the one construction in which "you don't need this" costs the client nothing.
The positioning is derived from market analysis, not assumed: the mass segment is saturated (~4,200 salons in Kyiv, ~71,600 beauty sole traders nationwide), the premium segment fell −38% in 2022 and recovered with double-digit growth — a durable premium demand. S
Google Trends (5 years, Ukraine) reinforces the position: category queries are falling — "cosmetology" −60%, "makeup" −60%, "laser hair removal" −50% — while "botox" +10% and "plastic surgery" doubles. Mass beauty search is leaving, a considered medical procedure is stable. Discovery has moved to Instagram: competitors hold 200–245K followers. S grade B
| Network | Branches | IG followers | Segment / tone |
|---|---|---|---|
| Lazerhouse | 75 | 207K | premium-mass · promo, deals |
| Lumenis | 27–32 | 245K | premium-clinical · educational content |
| TC Laser | 45 | — | mid |
| Slim | 4 | — | premium — a direct LUMÉ analogue |
| Mediskin | 9 | — | mid |
Fig. 1. Real device-cosmetology networks in Ukraine (uba.top + Instagram, 06.09.2026). LUMÉ at 4–5 studios in Kyiv = a mid-premium player, in Slim's league. Competitors are 17–18 years older and larger — LUMÉ (a boutique) realistically 15–40K followers D. LUMÉ's tone is closer to Lumenis's clinical than to Lazerhouse's promo.
Macro — real numbers, grade A. The LUMÉ trajectory — a simulation M. Full version — research/market_retrospective.md.
| Year | Real GDP | Inflation | Real wage | $ rate | Cosmetology market |
|---|---|---|---|---|---|
| 2021 | growth | +10.0% | — | 27.4 | peak, dense competition |
| 2022 | −28.8% | +26.6% | −11.9% | 32.7 | collapse → V-rebound in ~1 month |
| 2023 | +5.5% | +5.1% | +3.7% | 36.7 | +15% on −30% supply → demand ×7 |
| 2024 | +2.9% | +12.0% | +14.4% | 40.2 | ~10–15% |
| 2025 | +3.6% | +8.0% | +6.2% | 41.7 | prices +33.6% (Kyiv +20.3%) |
| 2026 H2 | slowing | slowing | — | 44.5 | "beauty boom", tightening regulation |
Fig. 2. LUMÉ revenue 2019–2026, a simulation M. Studio 1: 2 rooms in 2019 → 4 by 2023; No. 2 — Q4 2023 (into the "demand ≫ supply ×7" window); Nos. 3–4 — 2025 on the price wave; No. 5 — Q3 2026. ₴3.5M (2019) → ₴67M (2026).
The key finding: beauty services rose +33.6% in 2025 (Barb.ua; Kyiv +20.3%) — 3–4 times faster than CPI. The sector passes rising costs into prices with no loss of demand. In the model LUMÉ's prices are indexed by service category, not by inflation.
Studio 1 survived 2022 on a single location (₴4.5M, −21% on 2021): a generator in autumn 2022 (~₴150K) paid for itself with winter occupancy while only 15–20% of salons were working. The 2022→2023 jump (×2.5) is demand recovery (×1.7) + an added room + a higher ticket; M, but plausible for a surviving player in a year of supply shortage.
Everything is computed from _bible/model/params.py → lume_bible_model.xlsx. LTV/CAC — on gross profit (the same method as movement VII).
| Service line | Ticket ₴ | Materials ₴ | Contribution % | Share of revenue |
|---|---|---|---|---|
| Injectables (botox, fillers, PDO threads, biorevitalisation) | 8,000 | 3,360 | 48% | 57% |
| Device facial (microneedling RF / laser / SMAS) | 4,600 | 230 | 81% | 24% |
| Care protocols and peels | 2,100 | 420 | 68% | 14% |
| Laser hair removal | 1,300 | 65 | 81% | 4% |
| Diagnostics and consultations | 900 | 18 | 80% | 2% |
| Weighted | 4,228 | — | 60% | 100% |
Tickets cross-checked against Slim's full price list (a LUMÉ analogue): botox per zone ₴3.1–4K, Full Face ₴15K, fillers/Sculptra/GANA ₴8–22K, PDO threads ₴3.5–12.5K, microneedling RF ₴6–8.5K. Injectables — 48% of revenue, materials 44% (Sculptra/GANA — vial cost). The mix M drives the whole economics. LUMÉ doesn't run body-LPG/Endosphere.
| P&L for one studio · Podil · 75% utilisation · H2 2026 | ₴/mo |
|---|---|
| Visits per month | 360 |
| Revenue (ex-VAT) | 1,449,360 |
| Materials / products | −380,149 |
| Practitioners — base (sole traders: 3 cosmetologists + 2 doctors) | −242,000 |
| Practitioners — bonus on service value | −180,094 |
| Rent | −125,590 |
| Payroll, employees (2 admins, a nurse, a manager) | −124,000 |
| Social contribution 22% on employees | −27,280 |
| Other fixed (utilities, software, medical waste, insurance, generator) | −87,200 |
| Marketing (6.0% of revenue) | −91,325 |
| Equipment depreciation + lease (instalments) | −13,750 / −24,750 |
| EBITDA | 179,394 |
| Profit tax 18% | −29,816 |
| Net profit | 135,828 · 9.0% |
Fig. 3. Check: sources give an owner's net margin of 10–15% (rented) / ~20% (own premises). LUMÉ's ~9% for a mature studio is slightly lower: some equipment is on instalments, the rent for Pechersk/Centre is estimated above the office corridor (M, grade C — live listings give $17–23 for an office / $25–31 for street-retail). The mature Obolon studio — 10.5%.
Fig. 4. Client unit economics — a two-segment cohort: 42% of new clients become regulars (first-visit retention S), year-on-year retention 70% S, a 5-year horizon. CAC by channel: Meta ₴1,942, Google ₴1,504, influencers ₴1,723. The first visit is modelled as a trial (consultation + a light procedure, check ≈₴2,300 M, weights synced with the Takt VIII dataset), not as an average ₴4,220 visit — so it covers ~56% of CAC, not 103% as in v0.5. The non-resident advertising tax (art. 141.4.6) no longer exists — repealed by Law No. 1525-IX effective 01.01.2022; v0.5 wrongly kept it active.
| Demand: built up from the districts of presence · [D] on top of [S] | people |
|---|---|
| Women 25–55 in Kyiv (TAM, actual population 3.32M) | 766,920 |
| Population of the 5 districts of presence (actual, ×1.125 on registered) | 1,361,763 |
| Women 25–55 in the catchment (incl. +25% inflow from other districts) | 419,423 |
| Of those, able to afford regular medical cosmetology | 54,525 |
| Unique LUMÉ clients/year · of them regulars | 4,179 · ~1,755 |
Fig. 5. LUMÉ = 7.7% of the affordable segment in its catchment. Districts (Ukrstat 2016 C): Pechersk 152K, Podil 198K, Obolon 315K, Darnytsia 315K, Shevchenkivskyi 230K — 1.21M registered, ×1.125 to actual (3.32M de-facto per IOM DTM Round 24 vs 2.95M official). Women 55% B, aged 25–55 ~42% M. Able to afford ~13% M — RUN 12: v0.5 had 20%, derived from a "₴39.2K median", but that was Kyiv's average wage for March 2025, not the median. The real median is ~₴29–34K gross → net below the ₴25–31K threshold. IOM: 38% of Kyivans below the per-capita subsistence minimum (₴10,275), 35% cut medical spending. ⚠️ Ukrstat does not publish the income distribution openly — the parameter stays [M].
Fig. 5-b. Cross-check from the supply side. Kyiv has 370 cosmetology clinics (372 centres / 365 organisations doing injectables — Barb.ua S) ≈ 925 rooms ≈ 1.22M visits a year. LUMÉ accounts for 1.39% of city visits on 2.16% of the rooms — it takes a smaller share of flow than of capacity. The "premium multiple" (client share ÷ capacity share) = 3.5× — RUN 12 raised it from 2.55× (via an honestly smaller affordable segment) above the 1.5–2.5 norm. The model catches this itself: the network's financial plan (4,179 clients) sits at the optimistic edge of the honestly-sized Kyiv premium demand. The main conclusion "we hit capacity" is unchanged — a second boundary is now visible. The check appeared in v0.5 (v0.3 gave LUMÉ 21% of the market — implausible against a fragmented market where the top player holds ~5% of Kyiv locations).
| Network, 2026 (monthly, ramp + seasonality) | Revenue ₴ | Net ₴ | Margin |
|---|---|---|---|
| Pechersk | 18,264,960 | 1,370,816 | 7.5% |
| Podil | 18,264,960 | 1,673,073 | 9.2% |
| Obolon | 18,223,407 | 1,949,262 | 10.7% |
| Pozniaky | 15,683,969 | 701,361 | 4.5% |
| Centre (opened Q3) | 1,465,611 | −1,256,966 | start-up loss |
| TOTAL | 71,902,907 | 4,437,547 | 6.2% |
Fig. 6. Network revenue ≈ $1.62M/year, ~4,179 unique clients. A network of 4 mature studios (a model check): 9.5% margin.
An LLC + a VAT payer + an MoH licence. Derived, not chosen. The network's turnover is clearly > ₴1M → the simplified tax regime is impossible. "A network of separate sole traders" = a textbook sign of business splitting, which the tax office hits en masse in 2025–26. S grade A
An MoH licence is mandatory for injectables, laser, device treatments, medical peels (MoH order No. 117). State duty ₴3,328; setup with consultants 3–6 months. Doctors and cosmetologists are group-3 sole traders (avoiding the social contribution); admins, the nurse, the manager are employees.
Medical-service advertising (Advertising Law art. 21): featuring doctors in ads, fear-based framing, "before/after," "the most effective" are prohibited. This coincides with the "trust instead of fear" position.
The non-resident advertising tax no longer exists. Sub-clause 141.4.6 of the Tax Code (20% on Meta/Google payments "from own funds") was removed by Law No. 1525-IX of 03.06.2021, effective 01.01.2022 — the same law that introduced the "Google tax" (VAT on e-services). v0.5 of the model wrongly kept it active — its status had been frozen at the "bill passed a first reading" stage. Only 20% VAT on non-resident advertising applies (reverse-charge, art. 208): for a VAT-paying LLC with taxable activity, output = input → net on the media budget is 0%. S grade A (text of Law No. 1525-IX, zakon.rada).
Studio opening capex: ≈ ₴3.6M (within the ₴1.5–5M corridor for medical cosmetology). Equipment ₴1.5M (a diode laser ₴450–940K + a SMAS Ultraformer ₴0.8–1.5M + RF), a medical-facility fit-out ₴650K, a starting product stock ₴450K, the licence ₴300K. ~45% of the equipment is on manufacturer instalments, the payment a separate P&L line.
The booking system is Ukrainian (EasyWeek / Appointer), not Altegio. Altegio (ex-YClients) is tied to the Russian VK holding — a Forbes investigation (2023, confirmed in 2026), SBU reports of client-data leaks. For a network where the client record = medical data and the position is "trust," Russian booking software is a direct reputational and security risk. S grade A/B
The automation is built: 5 n8n workflows (107 nodes) + 2 Dify bots + a single data layer — lead intake with attribution and SLA, message routing with a classifier, deadline control, campaign guardrails, synthetic form monitoring. They all pull parameters from this Bible. Automation overview, guide — automation/SETUP.md.
provider_cabinets_avg 2.5 and provider_visits_cab_month 110 have no external figures (indirect support: 2.7 specialists/centre; 60–165 visits/workplace). The cross-check (LUMÉ 1.39% of city visits) inherits this grade C.premium_income_share = 0.13 depends on itWhat is fully incorporated: retention benchmarks set the LTV (a two-segment cohort); equipment leasing — a P&L line; Kyiv demographics — demand built up from the actual districts of presence, cross-checked against the supply side (370 clinics); the regular-client ticket cross-checked against a source; Google Trends — seasonality and the discovery shift; competitors' Instagram data — the competitor map; UA CPL 2025 — grade C (the one UA anchor: Rabbit Marketing 200–600 ₴/lead).