# RealFriend Financial Model AI-powered chat assistant that helps users find rental and for-sale apartments, replacing the traditional real estate broker relationship with a conversational AI agent. - Canonical: https://finamodel.com/startups/realfriend - Excel download: https://finamodel.com/startup-models/realfriend.xlsx - Category: AI/ML - Model type: Marketplace / GMV - Funding round: Seed - Funding: $4.4M - Founded: 2020 - Geography: Two active markets - Tel Aviv (launched 2018) and New York City (beta launched March 2019). [DECK slides 8, 9] - Customer: B2B ## About the company Realfriend's conversational AI assistant, Luke, helps renters and buyers search listings, capture preferences, schedule viewings, and receive follow-up suggestions. The hybrid service replaces parts of the traditional broker relationship with 24/7 chat and human hand-off where necessary. The marketplace earns 25% of an agent's commission on a closed referral, not a subscription. It reached 53,800 engaged users in Tel Aviv during 2019 and 2,500 monthly active users in New York, where engagement was growing 25% month on month. The model tracks users through conversations, referrals, closed transactions, and agent-commission revenue. It tests activation, retention, referral conversion, transaction value, consumer acquisition cost, agent supply, support and product cost, gross margin, monthly burn, and runway. ## What's included - 5-year monthly revenue build with stage-appropriate growth assumptions - Full P&L, headcount plan, and operating-expense schedule - Cash-flow statement, runway, and burn-rate tracking - Valuation via exit multiple with a DCF cross-check - Returns analysis with MOIC and IRR - Unit economics including CAC, LTV, payback, and cohort retention ## Product & value proposition - Conversational AI assistant ("Luke") available 24/7 via chat that searches, filters, and schedules apartment viewings on behalf of users. - Hybrid AI: bot handles search and scheduling; human hand-off implied ("hybrid AI real estate assistant"). - User flow: onboarding via chat → natural-language preference capture → proactive listing suggestions → in-app showing scheduling → post-visit follow-up with alternatives. - Longer-term vision: expand beyond real estate to loans, jobs, cars - AI advisor for all major financial decisions. ## Market - US real estate commissions pool: $100B - Referral fees sub-pool (25% of commissions): $10B - US millennials = 40% of home buyers - Tel Aviv rental market: ~50K apartments rented per year - No formal TAM/SAM/SOM breakdown or sourced market-size methodology in deck. ## Revenue model - Single revenue line: 25% referral fee from real estate agents on successful closings. - No subscription, SaaS, or advertising revenue mentioned. - No pricing per transaction stated explicitly; implied that agent pays 25% of their commission to Realfriend on each deal closed. - Channel: direct-to-consumer acquisition (word-of-mouth referenced for Tel Aviv growth) → agents pay referral on conversion. ## Traction & metrics - Tel Aviv: Launched 2018; 53,800 engaged users in 2019; "quickly became market leader in rentals market"; grew word-of-mouth only. - New York: Beta launched March 2019; 2,500 Monthly Active Users (as of Feb 2020). - Engagement growth: 25% month-over-month growth in engagement (NYC). - Retention / satisfaction: >60% of users would be "very disappointed" if service disappeared (Sean Ellis test). - NPS: NYC = 40 (after 1 year); Tel Aviv = 65 (after 2 years). Benchmarked vs. Zillow (-8), AI assistants (30–40), Tesla/Apple/Airbnb (65+). - No revenue figures, no transaction volume, no closed-deal counts disclosed. ## Competition / moat - Zillow positioned as an inferior alternative (NPS -8 vs. Realfriend's 40–65). - AI assistants (Hey Google, Amazon Alexa) cited as benchmark (NPS 30–40). - Moat framing: trust/brand ("world class trusted brand"), conversational relationship depth ("Your mom introduces him"), word-of-mouth organic growth. - No formal competitive landscape slide. ## Team & funding ask / use of funds - Two co-founders: Luke (NYC) and Dooron (Tel Aviv). - Background: working together 15 years, served in IDF special forces, launched multiple businesses together. - No funding ask amount, valuation, or use-of-funds breakdown disclosed in deck. --- ## Recommended financial model - **Archetype + why:** Marketplace / referral-fee revenue model with user-funnel P&L. Realfriend earns a 25% cut of agent commissions on closed deals; the core model should track users → conversations → referrals → closed transactions → referral revenue. Closest archetype is a marketplace/lead-gen model, not SaaS (no subscription revenue) and not a 3-statement operating forecast at this stage given the early traction. - **Forecast horizon & granularity:** 3-year monthly model (2020–2022), consolidating NYC and Tel Aviv as two separate market tracks that roll up to a combined P&L. - **Key drivers & assumptions:** *Demand / user growth* - NYC MAU at model start: 2,500 - NYC MoM engagement growth: 25% - will taper: use 25% in H1 2020, declining to 8% by end of year 2 as market saturates - Tel Aviv engaged users 2019: 53,800 - model as ~4,500/month average active (53.8K/12); MoM growth 5% given mature market - New market expansion (e.g. LA, Chicago): 1 new city per year from 2021 onward, ramping like NYC beta curve with 6-month lag *Conversion funnel* - User-to-referral conversion (users who successfully close a deal via Realfriend): 3%–5% per cohort-month, consistent with consumer real estate search intent - Average US rental deal commission: ~$2,000–$3,000 (one month's rent as broker fee in NYC); sale deal commission: 2.5%–3% of median NYC price (~$750K) = ~$19K–22K per sale - Realfriend take: 25% of agent commission - Mix: 80% rentals / 20% sales to start; shift to 60/40 over 3 years as brand builds *Revenue per closed deal* - Rental referral revenue per deal: $500–$750 (25% of one-month broker fee) - Sale referral revenue per deal: $4,750–$5,500 (25% of ~2.5% commission on ~$750K median) *Costs* - Headcount: 6 FTEs at start (2 founders + 4 engineers/ops); scaling to 20 by end of year 2 - AI/infra costs: $5K/month at 2,500 MAU, scaling linearly with users - Sales/partnerships (agent network): 1 BD hire per city per year - No inventory, no balance sheet complexity - asset-light - **Scenarios (Base / Bull / Bear - which variables flex):** - Base: 25% MoM NYC user growth tapers to 8% by Y2; 3% conversion; 80/20 rental-sale mix - Bull: Growth sustains at 20% MoM through Y1; 5% conversion; faster city expansion (2 new cities in 2021) - Bear: Growth slows to 10% MoM by Q3 2020 (COVID impact plausible given Feb 2020 deck date); 2% conversion; no new city until 2022 - **Required sheets / outputs:** 1. Assumptions & toggles (centralized) 2. User funnel (MAU by city → referrals generated → closings) 3. Revenue build (rental referrals + sale referrals by city, monthly) 4. OpEx build (headcount plan, infra, marketing) 5. P&L summary (gross revenue, gross margin, EBITDA) 6. Cash runway (monthly burn vs. funding; key for seed raise) 7. Scenario comparison table (Base / Bull / Bear on revenue and runway) ## Frequently asked questions ### Is the RealFriend financial model free? Yes. The RealFriend model is a free Excel download with live formulas. ### Can I change the assumptions? Yes. The workbook is editable and its live formulas recalculate when assumptions change.