SB StartupBasket
All ideas
79 /100 GO Low complexity

DealKaro — WhatsApp AI sales assistant for India's solo real estate brokers

WhatsApp AI that qualifies leads, books site visits, and follows up for India's solo real estate brokers.

PropTech SaaS India WhatsApp-firstAI-agentSMBMultilingualUPI-native
— views
Evaluation Scores
79/100

GO

Overall Score

17
Problem
12
Demand
14
Build
12
Distrib.
11
Revenue
8
Time
5
Defense

DealKaro

1. One-liner

WhatsApp AI that qualifies leads, books site visits, and follows up for India’s solo real estate brokers.

2. Trend signal — why now?

India has roughly 1 million independent real estate brokers — mostly solo operators running their business from a smartphone and WhatsApp. They’re the backbone of a $4 billion brokerage fee market inside a $23 billion+ real estate industry. And 85% of them don’t use any technology beyond WhatsApp itself.

Here’s the thing: 48% of buyer leads never get a response. Not because the broker doesn’t care — because the broker is standing in a half-built flat in Noida showing a 2BHK to another buyer when three 99acres leads hit their phone. By the time they check WhatsApp, the buyer called two other brokers.

Three things changed in the last 12 months that make this idea ripe now:

  1. WhatsApp Business API got dramatically cheaper in India. Meta’s 2026 pricing puts utility messages at ₹0.11 and service replies inside the 24-hour window at zero. That makes a high-volume conversational AI product economically viable at Indian price points for the first time.

  2. AI inference costs collapsed. You can now run a multilingual, context-aware conversational agent (Hindi + English + regional language understanding) for fractions of a cent per interaction. A year ago, the unit economics of a ₹499/month product with heavy AI usage didn’t work. Now they do.

  3. India’s proptech sector is booming but ignoring solo brokers. $550M was invested in Indian proptech in 2025 — but almost all of it went to full-stack platforms (Square Yards, Infra.Market), marketplace apps (NoBroker, Housing.com), or enterprise CRMs (Sell.Do at ₹20K-60K/month). Nobody is building for the solo broker with a ₹500/month budget.

Provenance:

3. The opportunity

The gap is absurdly clear: India’s independent real estate broker — the person who actually sells 80% of residential property in the country — is running a ₹5-20 lakh/year brokerage business on WhatsApp forwards and memory. They receive leads from 99acres, MagicBricks, Housing.com, and walk-ins. These leads arrive as phone calls, portal notifications, and WhatsApp messages. The broker juggles 20-30 active inquiries at any time, with exactly zero system to track who’s been contacted, who needs a follow-up, and who is ready for a site visit.

The incumbent solutions are all wrong for this buyer:

  • Sell.Do, LeadSquared: Enterprise CRMs at ₹20K-60K/month. Designed for developers with 50-person sales teams. A solo broker won’t even get through the onboarding.
  • TeleCRM, PropertyCRM: Better priced (₹250-500/month) but still desktop-first dashboards that require data entry. The broker’s actual workspace is WhatsApp — not a browser tab they’ll never open.
  • Gallabox, Spur, opZynic (chatbot platforms): These sell to developers and large brokerages who have IT staff to configure chatbot flows. A solo broker can’t set up webhook integrations.

The opportunity: build the CRM that lives inside WhatsApp. Not a dashboard that integrates with WhatsApp — a product where WhatsApp is the interface. The broker forwards a portal lead to a WhatsApp number, and the AI does everything: responds to the buyer, asks budget/location/BHK preferences, shares property photos, schedules site visits, sends reminders, follows up after 3 days, and gives the broker a daily “here are your hot leads” summary. Zero app downloads, zero logins, zero training.

4. Target market

  • Primary customer: Independent/solo real estate brokers and small 2-3 person brokerages in Indian Tier-1 and Tier-2 cities. Typically male, 25-45 years old, smartphone-first, Hindi/regional language speaking, handling residential sales and rentals. Monthly income ₹30K-1.5L.
  • Why they buy: They’re losing deals to faster brokers. They know they’re forgetting to follow up. They’ve seen leads go cold because they responded 6 hours late. Every lost deal = ₹20K-50K in commission gone. A tool that saves even 2 deals a month pays for itself 40x over.
  • Rough TAM reasoning: ~1 million independent brokers in India. Even at 2% penetration = 20,000 paying customers × ₹500/month = ₹1.2Cr MRR ($144K MRR). At 10% penetration = ₹6Cr MRR ($720K MRR). The addressable slice of brokers who handle 10+ leads/month and earn enough to pay ₹500/month is roughly 300K-400K.
  • Why now for them: Portal costs are rising (99acres premium packages went up again), lead quality from portals is declining (users report 90% of MagicBricks leads haven’t shown interest in the specific property), and competition among brokers is intensifying. They need to convert faster or lose to the next broker in the buyer’s WhatsApp inbox.

5. Product sketch (MVP)

  • Lead capture via WhatsApp forwarding: Broker forwards a 99acres/MagicBricks lead notification (or a buyer’s first message) to DealKaro’s WhatsApp number. AI extracts the buyer’s details.
  • Auto-qualification conversation: AI initiates a WhatsApp conversation with the buyer — asks budget range, preferred localities, BHK configuration, timeline, and whether they’re an investor or end-user. Multilingual (Hindi + English code-switching, which is how Indian real estate conversations actually happen).
  • Property matching & sharing: Broker pre-loads their inventory (photos + details via WhatsApp itself — send photos, type “3BHK Sector 77 Noida 85L”). AI matches buyer preferences to available properties and shares curated options.
  • Site visit scheduling: AI proposes available time slots, confirms with the buyer, sends the broker a calendar summary via WhatsApp, and sends a Google Maps pin + reminder to the buyer 2 hours before.
  • Automated follow-up sequences: If a buyer goes quiet after inquiry, AI sends a tasteful follow-up after 2 days, then 5 days, then 10 days. Broker can customize the cadence.
  • Daily lead summary: Every morning at 9 AM, the broker gets a WhatsApp message: “You have 4 hot leads (responded in last 24h), 7 warm leads (last contacted 3-5 days ago), 12 cold leads. [Buyer name] is ready for a site visit — call now?”
  • No app, no dashboard, no login. The entire product lives inside WhatsApp. A basic web dashboard exists for inventory management, but it’s optional.

6. AI angle — what’s load-bearing

Remove the AI and this product ceases to exist. The entire value proposition is an AI agent that:

  1. Reads and understands messy lead data — portal notifications come in varied formats (99acres email alerts, MagicBricks SMS, Housing.com push notifications). The AI parses these into structured buyer profiles.
  2. Conducts natural Hinglish conversations — Indian real estate conversations are a mix of Hindi and English, with property jargon, locality names, and pricing conventions (₹85L, 85 lakhs, eighty-five lacs). The AI needs to handle this natively.
  3. Qualifies leads with judgment — not just filling a form, but understanding that a buyer asking “koi 3BHK near metro under 80L?” is more serious than someone asking “Noida mein kya rates chal raha hai?” and prioritizing accordingly.
  4. Times follow-ups intelligently — knowing that a buyer who visited a site on Saturday but didn’t call back by Tuesday needs a nudge, while one who said “I’ll decide after Diwali” should be contacted in November.

Without AI, this is just a WhatsApp broadcast tool — a commodity. With AI, it’s a virtual sales assistant that never forgets, never sleeps, and never loses a lead in an unread message pile.

7. Localization angle

This is an India-first play, and localization is the moat, not a feature:

  • Language: Real estate conversations in India happen in Hinglish (Hindi-English code-switching), with heavy use of local property terms (registry, possession, OC, carpet area, super built-up). A global chatbot trained on English real estate jargon would fail instantly.
  • Payment rails: UPI-based subscription payments. ₹499/month auto-debit via UPI Autopay — the broker doesn’t need a credit card, which most don’t have.
  • Pricing: ₹499/month (₹5,988/year) is roughly $60/year. This is 1/10th of what a US real estate CRM charges and well within the willingness-to-pay of a broker earning ₹50K-1.5L/month.
  • Distribution: WhatsApp groups are the primary social network for Indian real estate brokers. There are 330+ active real estate WhatsApp groups where brokers share listings, gossip, and tool recommendations. Word-of-mouth in these groups is the distribution channel.
  • Regulatory context: RERA registration is increasingly mandatory for brokers, creating a natural database of verified brokers to target.

8. Business model — path to $1M–$5M ARR

  • Pricing: ₹499/month (Starter — up to 50 leads/month), ₹999/month (Pro — unlimited leads, priority follow-ups, multi-language), ₹1,999/month (Team — 3 agents, shared inventory, analytics).
  • ACV: Blended ACV of ~₹8,000/year ($96/year) assuming 60% Starter, 30% Pro, 10% Team.
  • Rough math to $1M ARR ($83K MRR): 10,400 paying customers × ₹8,000/year = ₹8.3Cr = ~$1M. At ₹499/month average that’s roughly 14,000 customers if mostly Starter tier.
  • Rough math to $5M ARR: 52,000 paying customers, or 25,000 customers with higher ARPU from Pro/Team adoption + upsells. At 5% penetration of the 1M broker market, this is achievable.
  • Expansion path: (1) Per-lead pricing for high-volume brokers, (2) Developer partnerships — builders pay DealKaro to distribute leads to their channel partner network with built-in qualification, (3) Premium features: AI-generated property descriptions, virtual tour links, buyer intent scoring.

9. Go-to-market wedge — first 100 customers

  1. RERA broker directory scrape + personalized WhatsApp outreach. Every state RERA authority publishes registered broker directories. Scrape Haryana, UP, Maharashtra, and Karnataka RERA databases (~50K brokers). Send personalized WhatsApp messages: “Hi [name], I see you’re registered in [city]. We built an AI assistant that follows up with your 99acres leads while you’re on site visits. Try it free for 14 days — just forward your next lead to this number.” Target: 5,000 messages → 10% reply → 500 trials → 20% convert = 100 paying customers in 6 weeks.

  2. Real estate WhatsApp group infiltration. Join 20-30 active broker WhatsApp groups. Share a 60-second screen recording showing DealKaro qualifying a lead in real-time. Brokers in these groups actively share tools. Target: 50 trials from group referrals.

  3. Partner with 3-5 popular real estate YouTube channels. India has a thriving ecosystem of Hindi-language real estate YouTubers (10K-100K subscribers) who teach brokers how to use technology. Sponsor 3 videos with demo walkthroughs. Target: 100+ trials per video.

  4. 99acres/MagicBricks seller community forums. Post case studies showing “I was responding to leads 4 hours late, now my AI responds in 30 seconds.” Include before/after conversion numbers.

  5. On-ground activation in Noida/Gurugram broker clusters. Real estate brokers in NCR congregate in known office clusters. Send a field rep with a tablet to demo DealKaro in person. 1-hour demos → instant WhatsApp setup → same-day activation. Target: 10 brokers/day × 20 days = 200 trials.

10. Build complexity — justification

Low. The core product is a WhatsApp Business API integration (well-documented, multiple BSPs available like Gupshup, Wati, AiSensy) + an LLM-powered conversational agent (GPT-4o-mini or Claude Haiku for Hinglish conversations at <$0.01/conversation) + a lightweight database for lead tracking. No mobile app, no complex frontend. The optional web dashboard is a simple React/Next.js CRUD interface.

A solo technical founder can ship v1 in 4-6 weeks:

  • Week 1-2: WhatsApp Business API integration via BSP, lead parsing from forwarded messages
  • Week 3-4: Conversational AI agent (qualification flow, property matching, follow-up scheduling)
  • Week 5-6: Daily summary generation, UPI payment integration, onboarding flow

Total infrastructure cost at 1,000 customers: ~₹50K/month (WhatsApp API messages + AI inference + hosting).

11. Gating checklist

GatePass?Note
Legal in target market✅WhatsApp Business API use for customer communication is explicitly permitted. Real estate brokerage is legal and increasingly regulated (RERA). No data the product handles is classified as sensitive personal data under DPDP Act.
Ethical — no harm / dark patterns✅Product helps brokers respond faster to genuine buyers. No spam — all conversations initiated by buyer inquiry. Follow-up sequences respect opt-out.
Market exists (evidence above)✅1M+ brokers, $4B brokerage revenue, 48% lead response gap, multiple funded competitors validating the category.
1–5 person team can build this✅Solo founder can ship in 4-6 weeks. Low complexity — API integrations and conversational AI on existing infrastructure.
Launchable with <$50K / ₹40L✅Estimated launch cost: ₹3-5L ($3.5-6K). WhatsApp BSP setup + AI API credits + basic hosting.

12. Feasibility score

AxisWeightScoreNotes
Problem intensity2017/20Hair-on-fire for active brokers. 48% of leads get no response. Every missed lead = ₹20-50K lost commission. Brokers feel this pain daily.
Demand evidence1512/15Strong: 1M brokers, multiple CRM competitors exist (proving demand), documented conversion lifts from WhatsApp automation (3× improvement). Docked because no direct pre-orders yet.
Build feasibility1514/15Off-the-shelf WhatsApp Business API, proven BSPs, cheap AI inference. Solo founder, 4-6 weeks. The hardest part is Hinglish NLU tuning — but foundation models handle this well already.
Distribution clarity1512/15RERA directories are scrapeable, WhatsApp groups are accessible, YouTube creator partnerships are proven. But broker churn is high in real estate — retention requires demonstrating clear ROI quickly.
Revenue mechanics1511/15₹499-999/month is within budget for earning brokers. Unit economics work (WhatsApp API costs are low). But ARPU is very low in absolute terms — need volume to hit $1M ARR. Currency risk on costs (AI APIs priced in USD).
Time to first revenue108/1014-day free trial → paid conversion. First revenue within 6-8 weeks of launch. UPI Autopay makes payment frictionless.
Defensibility105/10Low moat initially — any WhatsApp chatbot provider could add real estate templates. Defensibility grows with: (1) Hinglish conversation quality improving with data, (2) broker inventory data creating a lightweight listing network, (3) builder partnerships for lead distribution.
Total10079/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy — This is an AI + API integration play. The founder needs to be comfortable with LLM prompt engineering, WhatsApp Business API, and conversational UX design. Sales-heaviness kicks in at scale (builder partnerships), but v1 is a pure technical build.

Key assumptions to validate (3–5)

  1. Assumption: Solo brokers earning ₹50K+/month will pay ₹499/month for a WhatsApp-based tool. How to test: Run a pre-launch landing page (WhatsApp click-to-chat) targeting 500 RERA-registered brokers in NCR. Measure sign-up rate for 14-day free trial. Need >8% conversion to trial.

  2. Assumption: Brokers will actually forward leads to the AI (behavior change). How to test: Onboard 20 brokers manually, give them white-glove setup. Track what % actively forward leads in week 2 vs week 1. Need >60% retention of forwarding behavior.

  3. Assumption: AI can conduct Hinglish real estate conversations at quality high enough that buyers don’t realize it’s a bot (or don’t care). How to test: Run 100 test conversations with real buyer inquiries. Measure buyer response rate and site visit booking rate vs broker’s manual baseline.

  4. Assumption: ₹499/month ARPU is sufficient for unit economics at scale. How to test: Model costs at 10K, 50K, 100K customers. WhatsApp API costs + AI inference must stay below 40% of revenue.

Risk flags

  1. [Platform dependency]: Entire product depends on WhatsApp Business API. Meta could change pricing, terms, or restrict AI-driven conversations. The January 2026 terms update already targeted “AI providers” — though task-based bots remain compliant, the regulatory surface is uncertain.

  2. [Low ARPU / high volume requirement]: At ~$96/year ARPU, reaching $1M ARR requires 10,000+ customers. This is a grind — not a “land 50 enterprise accounts” business. Churn management becomes critical.

  3. [Market timing]: If a well-funded player like Square Yards, NoBroker, or 99acres themselves launches a similar feature as a free add-on to their platform, the standalone product gets squeezed. Speed to market and community lock-in matter.

  4. [Broker churn]: Real estate brokerage in India has high turnover. Many brokers enter and exit the profession seasonally. Customer base could be unstable.

14. Structured verdict

Score:                  79/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical solo founder with basic Hindi proficiency, comfortable with LLM APIs and WhatsApp integrations. Ideally based in NCR/Mumbai for on-ground broker access.
Time to revenue:        6–8 weeks
Capital to launch:      ₹3–5 lakh ($3.5–6K)
Top 3 assumptions to validate first:
  1. Solo brokers will pay ₹499/month — test with 500 RERA-listed brokers via WhatsApp outreach, measure trial conversion >8%
  2. Brokers will forward leads consistently — onboard 20 brokers, track forwarding retention >60% in week 2
  3. AI Hinglish conversation quality is good enough — 100 test conversations, measure buyer response rate vs manual baseline
Kill criteria:
  - Abandon if <5% of trial users convert to paid after 14-day trial
  - Abandon if <40% of onboarded brokers are actively forwarding leads by week 3
  - Abandon if WhatsApp Business API terms change to prohibit AI-driven buyer conversations

15. Next step — 1-week validation sprint

  • Day 1: Set up a WhatsApp Business number via Gupshup/Wati. Build a basic lead-forwarding flow (broker forwards a message → AI parses it and sends back a confirmation). No fancy AI yet — use a structured template.
  • Day 2: Scrape 500 RERA-registered broker contacts from Haryana and UP RERA portals. Prepare a personalized WhatsApp message with a 60-second demo video.
  • Day 3: Send outreach to 500 brokers. Track: open rate, reply rate, “tell me more” rate.
  • Day 4: For every broker who replies positively, manually simulate the AI flow — you be the AI. Forward their test lead, respond as the AI would, qualify the buyer, propose a site visit. Track the broker’s reaction.
  • Day 5: Decide go/no-go based on: (a) Did >40 brokers respond positively to outreach (>8%)? (b) Did >10 brokers complete the simulated flow? (c) Did >5 brokers say “I would pay for this”?

The validation must produce a measurable signal. If fewer than 5 out of 500 brokers express willingness to pay, the idea is a PASS — the pain isn’t acute enough, or the channel doesn’t work. Move on.

Interested in a detailed proposal?

Get a deep-dive with market research, competitive analysis, and implementation roadmap.

Contact us

info@startupbasket.ai