TEUM — Investment memorandum
25 September 2026 · Concept-stage venture · Currency: KRW unless stated otherwise
The investment thesis
TEUM helps an independent beauty studio turn an empty appointment into a completed visit. A merchant adds an available time, reviews eligible returning customers, and approves an invitation. The customer books through a mobile link. The merchant sees completed-visit revenue and, once a controlled pilot is implemented, the incremental contribution created by the intervention.
The attractive starting point is a short, economically legible transaction: one available hour, one suitable customer, one observed visit. The merchant already has the customer relationship. The consumer does not need another app. The product can begin with one manually entered slot before negotiating a calendar integration. These are reasons to prioritize a rapid test, not evidence that adoption or retention has already occurred.
The proposed company starts with 1–3-person, nonmedical nail and lash studios in Seoul's Seongsu–Konkuk area. It expands only when a repeatable acquisition channel and a measurable return on the subscription are established. The initial buyer should have documented marketing permission, at least 50 reachable consenting customers, services around KRW40,000–100,000 and at least six eligible gaps per month. These are screening hypotheses.
Why this opportunity was selected
Five Korean opportunities were compared: beauty revenue recovery, restaurant margin software, academy workflow software, family care coordination, and medical travel operations. TEUM scored 84/100 under a rubric that puts 25% on acquisition speed and 25% on demand and willingness to pay. These scores are author judgments informed by category evidence, not customer research. Moving 15 percentage points from speed to recurrence makes restaurant software rank first, 82 versus TEUM's 81. The choice is sensitive to the user's goal of fast app validation.
Official evidence establishes business conditions, not product demand. Korea's Ministry of SMEs and Startups reported digital tool usage of 27.2% and competition pressure of 61.0% among surveyed small businesses in 11 major industries, using 2024 data published in March 2026. The survey does not establish nail-studio vacancy rates or TEUM willingness to pay. MSS source.
Existing paid beauty software and automated waitlists establish a real product category. They also create a difficult pricing test. Gongbiz displays a KRW19,800 monthly Standard list price and a KRW9,900 new-user offer. Fresha offers automated waitlist updates; its Independent plan displays USD19.95/month. TEUM's proposed KRW49,000 price must earn its premium through lower adoption effort and incremental profit, rather than a generic feature checklist. Gongbiz, Fresha pricing, Fresha waitlists. Prices accessed 25 September 2026; regional, tax and promotional conditions apply.
Product and workflow
The working local prototype presents a mobile app interface with four fixed bottom tabs, touch-based appointment cards, two-step invitation bottom sheets, and an in-app customer booking flow. It is demonstrated in a browser; no native binary or store release is claimed. It supports slot creation, overlap rejection, permission-aware customer eligibility, invitation review, a tokenized booking link, a single successful reservation under concurrent requests, opt-out, attendance confirmation, CSV import/export, settings and durable SQLite storage. The interface explains why a customer is eligible and keeps a 0% discount as the default.
The rule engine is deliberately inspectable: service preference, time since last visit, marketing consent and contact frequency. It does not use a trained prediction model or claim a probability of conversion. Learned ranking may be valuable after sufficient consented outcome data exists; an LLM is not required to test the core commercial proposition.
The prototype records attributed completed-visit revenue. It does not label that revenue as causal lift. Real messaging, payment collection, merchant identity, multi-tenant isolation and external calendar reconciliation are not connected. The dataset is synthetic. The locally working transaction demonstrates feasibility of the core flow, not production readiness.
Customer economics and pricing
Proposed Solo: KRW49,000 per month. Proposed Team: KRW99,000 per month, subject to testing the multi-staff value proposition. The prototype stores the plan selection but does not implement multi-staff scheduling. Prices exclude VAT. Usage-based messaging is assumed to be charged separately at cost in the initial model.
Illustration: six genuinely incremental completed visits × KRW60,000 × 70% contribution margin = KRW252,000. After KRW49,000 subscription and KRW5,000 messaging, the owner retains KRW198,000. Break-even is two incremental visits per month at those assumptions. Service contribution must include materials, variable labor and payment costs. Moving an already-planned visit forward is not automatically incremental revenue.
The first paid test offers comparable merchant groups KRW29,000, KRW49,000 or KRW79,000 after an observed value trial. Use actual payment and retention rather than a willingness-to-pay survey alone. Stop increasing product scope if price, support burden and measured lift cannot support a healthy subscription.
Distribution: the first 100 activations
The 90-day target is 100 pilot activations and 30 paying accounts. No pilots, partners or leads have yet been secured.
| Channel hypothesis | Funnel | Activations | Budget including acquisition labor |
|---|---|---|---|
| Local founder-led sales | 300 qualified shops → 120 conversations → 60 demos | 30 | 3,000,000 |
| Five educators or suppliers | 100 introductions × 40% activation | 40 | 3,000,000 |
| Activated-merchant referrals | 60 referral opportunities × 50% activation | 30 | 1,500,000 |
| Total | 30% activation-to-paid conversion | 100 / 30 paid | 7,500,000 |
This implies CAC of KRW250,000 per paying account, not KRW75,000 per free activation. All founder time, travel, referral credits and partner commissions need tracking. Partner compensation should be tied to retained paying merchants rather than unqualified introductions. A local route is an access hypothesis; introductions are not contractual distribution.
Market and global scale conditions
The appropriate denominator is qualified paying locations. A verified count of eligible Korean nail/lash studios was not established. The model therefore uses 10,000 / 20,000 / 40,000 assumed eligible locations at KRW49,000 monthly ARPA, giving annual opportunities of KRW5.88bn / 11.76bn / 23.52bn. These are sensitivity scenarios, not official TAM estimates. Active-license data, service mix, staff size, permission records and duplication must be checked before presenting a market count as fact.
Domestic expansion would move from nail/lash to other nonmedical appointment businesses only after demonstrating comparable economics. Japan is a candidate second market because the workflow can be tested at small appointment businesses; it is not a researched or committed expansion market. Before launch there, conduct 20 local merchant interviews, test channel economics, validate consent and consumer rules with local counsel, and secure one authorized calendar or industry partner.
A later English-language wellness product would face Fresha and other established tools directly. Acquisition must be won country by country; translation alone is not an expansion strategy. Fresha's self-reported 120,000+ businesses demonstrates category scale, not the number of accounts available to TEUM. Company source.
The venture-scale condition is explicit: 100,000 global paying locations at KRW149,000 monthly blended ARPA would produce KRW178.8bn ARR, approximately USD127.7m at an assumed planning rate of KRW1,400/USD. Neither that exchange rate nor those accounts are a forecast. Reaching that level requires richer multi-location value, embedded partner distribution and durable retention. A profitable local SaaS remains a different possible outcome if those conditions do not emerge.
Operating model
| Assumption | Downside | Base | Upside |
|---|---|---|---|
| Monthly ARPA | 39,000 | 49,000 | 69,000 |
| Gross margin | 65% | 80% | 85% |
| Fully loaded CAC | 400,000 | 250,000 | 180,000 |
| Monthly logo churn | 6% | 3% | 2% |
| CAC payback, months | 15.78 | 6.38 | 3.07 |
| Simple gross-profit LTV / CAC | 1.06 | 5.23 | 16.29 |
Simple LTV = ARPA × gross margin / monthly churn. It excludes discounting, cohort differences, reactivation and expansion and should not be treated as measured lifetime value. Base-case monthly churn implies only 69.4% annual logo retention. The downside largely eliminates attractive acquisition economics.
| Planning model, KRW million | Year 1 | Year 2 | Year 3 |
|---|---|---|---|
| Average paying accounts | 80 | 400 | 1,300 |
| Year-end paying accounts | 200 | 700 | 2,000 |
| Monthly blended ARPA, KRW | 49,000 | 59,000 | 69,000 |
| Recognized annual revenue | 47.04 | 283.2 | 1,076.4 |
| Year-end ARR | 117.6 | 495.6 | 1,656.0 |
| Gross margin | 70% | 78% | 82% |
| Operating expense | 420 | 600 | 900 |
| Operating result before financing and tax | −387.072 | −379.104 | −17.352 |
Average accounts, year-end accounts and ARPA are independent planning inputs here; no monthly cohort forecast has been validated. ARR is not recognized revenue. COGS includes estimated hosting, customer support and integration costs; pass-through messaging is excluded from both revenue and COGS. Additional hiring, partner fees, refunds or free messages may worsen the model. The annual plan is not a treasury cash-flow schedule.
Validation and funding proposal
Proposed pre-seed funding is KRW800m over 18 months: product/engineering/field operations KRW480m; acquisition and partnerships KRW120m; infrastructure/security/integrations KRW80m; legal/accounting/operations KRW40m; contingency KRW80m. This is an illustrative capital plan, not an existing financing offer. Revenue is not assumed to be guaranteed runway. Monthly cash planning and hiring quotes are required before setting a round size.
The initial operating roles are a product/commercial lead, an engineering lead and field/onboarding operations. Founder identities, biographies, relevant achievements and hiring commitments were not provided; none have been invented. Founder-market fit is an open diligence item.
Days 1–14: interview 20 merchants and inspect permissioned recent calendars. Days 15–45: activate 30 pilots and observe at least 300 eligible gaps, while testing price. Days 46–90: target 100 cumulative activations and 30 paid accounts. Mature 90-day paid retention can only be reported after each cohort has actually aged 90 days.
The production experiment should randomize comparable slots or business days within merchant/service/time strata. Customers competing for the same slot are not independent observations. Measure total visits and contribution over a follow-up window, account for displacement of existing bookings, and pre-register the required sample after observing baseline variance. Three hundred gaps is an observation target, not a power calculation. The prototype's customer holdout display is an interface demonstration only.
Scale gates: median incremental contribution of at least 3× the subscription; a suitably powered effect estimate whose lower confidence bound exceeds zero; paid conversion of at least 30%; mature 90-day paid retention of at least 85%; onboarding under 20 minutes; monthly support under 20 minutes per merchant; fully loaded CAC payback under nine months. Rework or stop if free incumbent workflows achieve the same result, incremental visits remain below two monthly, permissioned reach is too small, or repeated data entry erases the value.
Questions a serious investor should ask
Why will an incumbent not copy this? It can. Defensibility must come from distribution, low-friction authorized integrations and a superior measured operating outcome. None of these is an established moat today.
Why pay more than existing software? Only if a controlled merchant trial demonstrates incremental contribution comfortably above the fee. If an incumbent achieves equivalent results at lower total cost, the current thesis fails.
Does this replace Naver Booking? No replacement is required for the initial experiment. The prototype uses manual slots; future calendar writes require explicit partner authorization and conflict reconciliation. Manual dual entry must remain small enough to be sustainable.
Is this an AI company? The current product is a rules-based workflow. AI may later improve targeting or copy under merchant control. The business value does not depend on presenting deterministic rules as AI.
Can marketing invitations use Kakao Alimtalk? Promotional invitations and transactional booking confirmations must be classified separately. Alimtalk is an informational channel. Live deployment needs an appropriate provider, consent evidence, opt-out handling and applicable legal review. Kakao guide, statutory source.
What exists today? Research with primary-source links, a persistent local working prototype, test results, an editable investor deck, and a concrete experiment design. Customer demand, revenue, partnerships and venture-scale defensibility remain to be established.
Evidence access
The source register contains 14 primary or first-party sources, publication/observation dates and scope limitations: sources.json. The Korean report includes all five ideas, rubric, sensitivity and full customer-acquisition plan: research. Product screens, data model, API and production gates are in product specification.