Walk through the websites of the larger Indian real estate CRMs and you will notice something. The marketing talks about "brokers and developers" in the same breath. The demo is built around a developer's pipeline — units, towers, booking. The case studies are mostly Tier-1 developers with five-figure unit counts. Then, almost as an aside, there is a feature toggle for "broker mode".
Broker mode is rarely a thoughtful product. It is the developer schema with the inventory table renamed.
This is fine for a one-office, ten-agent firm that mostly works a single locality and lives inside WhatsApp anyway. It is not fine for a 25-agent broker firm with three offices, a resale book, a rentals book, and a sub-broker network. At that scale, the gap between what a packaged CRM does and what a broker firm actually does each week becomes the difference between a calm Monday morning and a fire.
This is a field guide for picking pipeline software if you are a broker firm with five to fifty agents in India. We will cover what the packaged products actually do, the four jobs of a real broker stack, where packaged CRMs end, and the thin custom layer that usually earns its keep.
What packaged broker CRMs actually solve
The Indian real estate CRM market has matured. Walk demos of Sell.Do, Leadrat, Smartx, LeadNXT, Realatic, Makanify, 91Leads, TeleCRM, Freshsales, and a dozen others and you will see roughly the same surface:
- Portal lead capture. Inbound from 99acres, MagicBricks, Housing.com, Facebook lead ads, Instagram, Google forms — all routed into a single inbox via official APIs or scraping.
- Lead deduplication. Phone-number based, sometimes with email and name fallback.
- Round-robin or rule-based assignment. By locality, by language, by team, by daily cap, by source.
- Pipeline stages. New / contacted / hot / site-visited / negotiating / closed / lost — usually customisable.
- Reminders. Follow-up nudges, missed-call alerts, daily activity reports.
- WhatsApp integration. Click-to-WhatsApp, template messages, basic two-way thread.
- Click-to-call and call recording. Increasingly with AI summary of the call — buyer budget, BHK, preferred area, urgency.
- Inventory list. A listings table — usually a denormalised view of properties with status, price, area, location.
- Reports. Funnel, source ROI, agent leaderboard, lost-reason analysis.
For a single-office, 5–15 agent firm working one or two micro-markets, this is almost enough. Pick one whose support team picks up the phone, pay the INR 1,200–3,000 per user per month, train the team, run it for a year. The pain you feel will be operational discipline, not software.
The schema starts to bend once you cross three thresholds at once — and most growing broker firms cross them in the same twelve months.
The four jobs of a broker pipeline stack
The mental model that has held up across every broker engagement we have looked at:
Job one — Inflow. Capture every lead from every source. Portals, social ads, your own website, walk-ins, reference calls, agent personal WhatsApp threads. Dedupe by phone. Attribute by source. Land it in one inbox that the firm — not an individual agent — owns.
Job two — Inventory. Maintain a live picture of inventory you can actually sell. Resale flats, rentals, builder allocations, off-market listings shared by other brokers. With freshness state — last verified, who verified, current ask, current owner urgency. This is the schema where packaged CRMs are weakest.
Job three — Matching. Take a buyer's requirement (3BHK in HSR Layout, 1.6–2 Cr, ready possession, family with school-age kids) and surface inventory that actually fits — and the visit logistics to show it. Most packaged products treat this as a manual filter on the listings table. Real broker firms do it as a workflow.
Job four — Commission ledger. Who brought this lead, who did the visits, who closed, who introduced the seller, what split was agreed, what has been invoiced, what has been collected. The unit economics of a broker firm live in this table. The packaged CRMs treat it as a free-text field in the deal record.
A packaged CRM solves job one well, job two partially, and jobs three and four poorly. That is the shape of the leverage.
Where packaged broker CRMs end
If you are a broker firm that has grown past 15 agents and two localities, here is where the demos start to feel thin.
Resale and rental inventory has no real schema. Most products treat inventory as a flat listings table — area, location, BHK, ask price. Real resale inventory needs: source (was this seller a walk-in, an owner-direct call, another broker's tip?), freshness (when did we last verify the unit is still available?), exclusivity (is this listed with us only, or with five other brokers?), owner urgency (how negotiable is the price?), photos and floor plan (because the buyer will ask), and a verification status. Without this, your agents will routinely show buyers units that are no longer available — the fastest way to lose trust in this business.
Matching is a filter, not a workflow. A buyer with a real requirement should produce an automated shortlist with viewing logistics — three units in HSR Layout that fit budget, sorted by what is actually showable this Saturday, with the seller's slot preferences already factored in. Packaged products give you a filter on the listings table and call it matching. The leverage is in the workflow: shortlist → owner-confirmation message → visit slot → buyer feedback after visit → re-shortlist. None of that happens in the listings filter.
Sub-brokers and the originator question. Once you have a sub-broker network — independent agents who bring you sellers or buyers in exchange for a share — the commission split becomes a graph, not a number. Lead came from sub-broker A, was worked by your agent B, closed with seller introduced by sub-broker C. The packaged CRMs do not model this; the firm rebuilds it in a Google Sheet, which is where the leakage starts.
Rental rent-roll is its own thing. If you run a rentals book, you also have a rent-roll problem — monthly rent collections, escalations, deposit refunds, brokerage trail commissions on renewals. Packaged sales CRMs do not handle this. Operators end up with either a separate property management product (which does not talk to the CRM) or another Google Sheet (which does not talk to anything).
Owner cockpit. As a broker firm owner, you want one screen — by team, by office, by locality — that tells you: how many leads, how many converted to visits, how many closed, what the brokerage value of the closed deals is, what is outstanding from clients, what is owed to sub-brokers. Most packaged products give you five separate reports that do not reconcile.
These are not edge cases. These are the jobs the firm actually does on a Wednesday afternoon.
The thin custom layer that earns its keep
We have written before about the three-layer stack for developers — and the pattern repeats here, with the roles slightly different. The packaged CRM solves the common 80%. The leverage sits in the seams, in a thin custom layer that talks to the packaged product underneath.
For a broker firm with 15–50 agents, the layer usually has four surfaces.
A unified lead inbox. WhatsApp Business API connection, portal lead webhooks, walk-in form, website form, missed-call capture — all landing in one inbox with dedupe and assignment. Most firms either run the packaged product's inbox (which misses WhatsApp inflow) or run agent personal WhatsApp threads (which misses everything else). The custom inbox closes both gaps. This is roughly the same surface we covered in the HVAC website + dispatch wedge — different vertical, same idea.
An inventory module with freshness state. A listings UI built around the verification workflow — last verified, who verified, photos, floor plan, owner availability slots, exclusivity flag, commission split agreed. Hidden from matching if not re-verified in N days. The point is not the table; the point is the weekly verification ritual the UI enforces.
A matching workflow. Buyer requirement → shortlist → owner-confirm → visit slot → buyer feedback → re-shortlist. Each step is a row in a database, not a chat message lost in someone's inbox. The agent runs this from a phone; the owner sees it in the cockpit.
Commission ledger and cockpit. Every deal records originator, working agent, closer, seller introducer, agreed split percentages, invoice raised, amount collected, sub-broker payout owed, sub-broker payout cleared. Reports roll up by office, by team, by locality, by month. This is the system of record for the firm's P&L. Building it as anything other than a custom layer — Google Sheet, packaged CRM free-text field, separate Tally entry — produces numbers that nobody trusts.
These four surfaces are the internal tool layer of the firm. They are also the four that nobody sells you off the shelf, because the workflow is too specific to your office structure, your locality mix, and your sub-broker rules. We have shipped variants of these on top of packaged products as a custom layer — same pattern as the F&B operators we built central-kitchen surfaces for.
What about WhatsApp?
In India, WhatsApp is not a channel — it is the medium. Every broker firm runs most of its actual work through WhatsApp, and yet most of the leads that arrive on WhatsApp die in personal threads. Industry estimates suggest 30–40% of portal leads never get a first contact because the routing breaks somewhere between the portal and the agent's phone.
A serious broker pipeline treats WhatsApp Business API as a first-class surface. Inbound leads from WhatsApp Business land in the same inbox as portal leads, with the same dedupe rules, the same assignment rules, the same SLA timer. Agent-side, the conversation thread is logged against the lead, not lost in the agent's personal phone. When the agent leaves the firm, the conversation history stays. This is one of the highest-ROI automation surfaces we have built in the vertical — and the build is small, two to three weeks for a competent team.
A worked example
A Pune broker firm with 28 agents across three offices — Baner, Wakad, Hinjewadi — running resale, fresh launches from three developer partners, and a small rentals book. Currently on a packaged CRM at INR 2,500 per user per month for portal capture and pipeline. Three Google Sheets behind the scenes: one for resale inventory, one for sub-broker payouts, one for rentals rent-roll. Owner spends Saturday morning reconciling.
A focused build looks roughly like:
- Phase 1 (8–10 weeks, INR 9–12L): Unified lead inbox (WhatsApp Business + portal webhooks), resale inventory module with freshness state, matching workflow with shortlist and visit-slot, owner cockpit reading from the packaged CRM and the new tables.
- Phase 2 (4–6 weeks, INR 5–7L): Commission ledger with originator/worker/closer/sub-broker split, payout reports, sub-broker login.
- Phase 3 (4 weeks, INR 3–4L) — optional: Rental rent-roll with collections, renewals, deposit refunds.
Plus a retainer of INR 45–60k per month for the slow trickle of changes — new portal added, new sub-broker commission structure, new locality, new report the owner wants. The retainer is the part most buyers under-budget for; we have written about how to think about post-launch retainers separately.
Total year-one cost: INR 22–28L on top of the packaged CRM seat licences. The packaged CRM keeps doing what it does well. The custom layer does what the packaged CRM cannot.
The honest answer
We get asked the binary version of this question — buy a broker CRM or build your own — and the honest answer for a 15–50 agent firm is almost always hybrid. Don't replace what works. Don't build what's commoditised. Buy the boring core: portal capture, dedupe, basic pipeline, WhatsApp click-to-message, call recording. Build the layer that's specific to your firm: unified inbox with WhatsApp Business, inventory with freshness, matching as a workflow, commission ledger, owner cockpit.
The market has fifty broker CRMs and very few of them are bad. They are just aimed at the common 80%. The 20% that decides whether a broker firm makes its number — inventory freshness, matching discipline, commission accuracy, owner visibility — is custom by nature, because it is exactly the workflow that makes your firm different from the firm down the road.
If you are evaluating broker software and the demo feels like a slightly-renamed developer product, that is not a coincidence. It is the shape of the market. The fix is not to keep evaluating; it is to buy the packaged core and build the thin layer that makes it yours.
If you are at that decision point and want a second opinion on what to buy and what to build, book a discovery call. We have shipped variants of this stack and can usually tell within thirty minutes whether the leverage in your firm sits in the inventory module, the commission ledger, or the WhatsApp inbox — and what a sensible first phase looks like.