AI for real estate agents: lead follow-up and site visits

AI can reply to property leads within minutes, ask about budget and location, follow up on schedule and book site visits, while you negotiate and show homes.

By the KutrimBots teamPublished 10 min read

AI helps a real estate agent most with the part that's hardest to do by hand: replying to every new property lead within minutes, asking the same qualifying questions (budget, location, configuration, when they want to move), following up on schedule and booking site visits into your calendar. You stay on the parts that need a person: calls with serious buyers, showing homes, negotiating and checking documents. In KutrimBots, an AI agent runs this with a built-in CRM, website chat, lead forms and a site-visit booking page, and asks for your OK before contacting anyone new.

This guide covers why leads go cold, what the AI should and shouldn't do, the setup, and the RERA points a person must still own.

Why do property leads go cold before you call them?

Because interest fades within hours, and many agents reply the next day or not at all. Research published in Harvard Business Review found that firms contacting a web lead within an hour were nearly seven times as likely to have a qualifying conversation as those that waited even one hour longer, and over 60 times as likely as firms that waited 24 hours or more. The same article's audit of 2,241 US companies found only 37% replied within an hour and 23% never replied at all, according to "The Short Life of Online Sales Leads".

Property leads have extra problems:

  • The same buyer enquires with several agents. Whoever replies first with useful options usually gets the site visit.
  • Leads pile up in many places: portal emails, ad forms, your website, calls you missed.
  • Ad leads expire. Meta keeps lead-ad form data downloadable for only 90 days after someone submits it, according to Meta's Business Help Center, so leads that never reach a CRM are simply lost.

What can AI do for a real estate agent?

AI takes the repetitive, time-sensitive work; you keep the relationship. Here's a fair split:

StageWhat the AI agent doesWhat you do
New enquiryReplies on your website chat at any hour; drafts a first email to form and portal leads for your OKApprove first messages with one tap
QualificationAsks budget, location, BHK, ready or under construction, loan status, who decidesRead the summary on each lead
MatchingSuggests listings from your inventory notes and price listDecide what to pitch
Follow-upKeeps a next step on every lead and follows up on timeTake the calls it reminds you about
Site visitsOffers free slots, books them, sends confirmationsShow the property
After the visitAsks whether the buyer came and plans the next step with youNegotiate
ReportingMorning brief: hot leads, overdue follow-ups, today's visitsRead it in two minutes

How does AI qualify a property lead?

It asks a short set of questions in a natural order and writes the answers on the lead, so you only call people who are ready. A good real estate script covers:

  1. Budget, for example "₹85 lakh to ₹1 crore", and whether that's all-in (registration, stamp duty, parking) or the base price.
  2. Location and configuration: 2 or 3 areas they'd consider, 2 BHK or 3 BHK, and a carpet area range such as 650 to 800 sq ft.
  3. Timeline: ready to move within 3 months, or fine with possession in 2028.
  4. Financing: own funds, a home loan already sanctioned, or still to apply.
  5. Who decides: a couple, parents, a company.
  6. Purpose: to live in, or to invest for rent.

In KutrimBots, the CRM & Leads plugin stores these as BANT notes (budget, authority, need, timeline), with a priority (hot, warm or cold), an expected value and a next follow-up date. You can rename the pipeline stages to match your work, for example New → Contacted → Visit booked → Visit done → Negotiation → Won or Lost, or keep separate pipelines for rentals and sales.

How do I set it up in KutrimBots?

Plan about an hour for the first setup.

  1. Hire the agent. Create an agent for sales and describe the job in plain words: the areas and projects you handle, price ranges, who you work with, and what it must never promise (prices you haven't confirmed, possession dates, loan approvals).
  2. Teach it your inventory. Add your current listings, price sheets, project RERA numbers and FAQs (parking, maintenance, possession) to Super Brain, so every reply uses the same facts.
  3. Put the AI where buyers arrive. Add the Chat Widget to your website or landing page with one line of code; the website agent answers questions from your help centre, takes contact details and offers site-visit slots. Visitors who share a phone or email become CRM leads. Our guide to adding an AI chatbot to your website shows the setup.
  4. Make a lead form. Start from the Forms plugin's lead template, ask for name, phone, budget, area and timeline, and point your ads and listing links at it. Each response creates a CRM lead and alerts the agent within a minute. Add a hidden utm_source field so you know which ad or listing sent each lead.
  5. Catch portal emails. If a portal emails you each enquiry, an automation can pass new enquiry emails from Gmail to the agent, which creates the lead and drafts a reply. The first email to each new person waits for your one-tap OK.
  6. Create the site-visit booking page. In Bookings, add a service such as "Site visit: Baner 2 BHK project", with a duration, the address, your visiting hours and the questions to ask. Share the link everywhere, including your WhatsApp Business greeting and away messages.
  7. Set the follow-up rhythm. The CRM wakes the agent when a follow-up is due, when a new lead hasn't been contacted, or when a lead stalls in a stage. A rhythm that suits property buyers:
WhenChannelWhat happens
Within 1 minuteWebsite chatThe agent answers and offers 3 to 5 visit slots
Within 15 minutesEmail, after your OK2 or 3 matching options and the booking link
Day 2EmailOne new listing within their budget
Day 5A call by youThe agent reminds you; you log the outcome
Day 10EmailPrice or possession update on what they liked
Every 30 daysEmailA short monthly update for cold leads, until they say stop

The CRM part works the same way for any business; CRM lead follow-up with AI agents explains that loop in detail.

How does AI book site visits?

Through a booking link with your real availability, so nobody waits for you to reply with times. Bookings shows free slots from your visiting hours, and your Calendars meetings block those times automatically, so a visit never lands on top of a registration appointment.

  • The buyer picks a slot from the link, or the agent offers 3 to 5 times in the website chat and books the one they choose.
  • Confirmations and reminders go out on WhatsApp from your business number, using message templates Meta has approved, set up once in Bookings → Settings.
  • Group visits work too. Set a group size for a project open house so several families can book the same Sunday slot.
  • No-shows are tracked. Mark who came; people who don't turn up get a friendly "missed you, book again" message, and you can check a buyer's booking history before you drive out.
  • After each visit, KutrimBots asks whether the buyer came, and the agent plans the next step with you: a price discussion call, a second visit with parents, or a polite close.

Every booking lands in your calendar, which you can subscribe to from Google, Apple or Outlook.

Example scenario

Example scenario: a two-person brokerage in Pune selling resale flats in Baner and Wakad and 3 new projects. It spends about ₹60,000 a month on lead ads and a portal plan and gets about 240 enquiries, around ₹250 per lead.

  • Before: enquiries are answered when someone gets to them, typically 4 to 6 hours later and next morning for anything after 8 PM. About a third are never contacted. Roughly 20 site visits happen a month.
  • With an AI agent: website and landing-page enquiries get an answer in the chat within a minute, at 11 PM too. Form and portal leads get a drafted first email that the owner approves from his phone, usually within 15 minutes. Every lead has a next step.
  • Qualification: of 240 enquiries, the agent marks about 70 as hot or warm with full notes, and the owner calls those the same day. The rest stay on a slow monthly follow-up instead of being forgotten.
  • Site visits: buyers book weekend slots from the link or in the chat, and WhatsApp reminders go out before each visit. Visits rise to about 35 a month in this example.
  • Time: the owner stops spending about 2 hours a day sorting enquiries and spends it on calls and visits.

These figures are illustrative assumptions, not results; what you see depends on your market, ad quality and how quickly you approve first messages.

What must you still do yourself?

AI should never stand in for you on trust, money or legal claims. Keep these with a person:

  • Calls and negotiation. KutrimBots' agent works by chat and email and reminds you when a call is due; you make the call and tell it what was agreed.
  • Showing the property and answering what you see on site.
  • Documents and money: token amounts, agreements, loan paperwork and registration.
  • RERA compliance. Real estate agents dealing in RERA-registered projects must register with their state's authority, as the Ministry of Housing and Urban Affairs explains. Section 10 of the RERA Act says a registered agent must not facilitate sales in unregistered projects, must keep prescribed records, and must not make false or misleading statements about the services, approvals or affiliations. Your AI's messages are your statements, so give it only verified facts and tell it to share each project's RERA number.
  • Advertising rules. Section 11(2) of the Act requires project advertisements to carry the RERA registration number and the authority's website. In Maharashtra, MahaRERA's Order 46A/2023 also made the project QR code mandatory on promotions from August 1, 2023. Check your state's rules before the agent drafts any ad or social post.

Two practical guardrails help: keep approval on for every first message to a new person, and write a "never say" list in the agent's instructions (guaranteed returns, possession dates the builder hasn't confirmed, "price will go up next week").

Getting started

KutrimBots gives you a team of always-on agents with the CRM, website chat, forms, bookings and calendar built in. Plans start at $20 a month for up to five agents, with 5,000 CRM leads and 1,000 website chats a month as of November 2026. You can try it free for 7 days: add a card or UPI, nothing is charged today, and it becomes the Solo plan after the trial unless you cancel.

Frequently asked questions

Can AI call my real estate leads for me?

Not in KutrimBots. The agent replies by website chat and email, reminds you when a call is due, and logs the notes you give it after the call. Some other tools offer AI voice calls; check consent and disclosure rules before using them.

Can AI follow up leads from property portals?

Yes, if the portal sends you each enquiry by email or you add the lead yourself. An automation passes new enquiry emails to the agent, which creates the CRM lead, drafts a reply for your OK and schedules the next follow-up.

What is the best AI for real estate agents?

Look for one tool that does four things together: answers enquiries fast, keeps every lead in a CRM with a next step, books site visits from your real calendar, and asks before contacting new people. A separate chatbot, CRM and calendar can work too, but the gaps between them are where leads get lost.

Is it allowed under RERA to use AI for property leads?

RERA regulates what you say and which projects you sell, not which tools you use. What the AI writes counts as your statement, so give it only verified project facts, include RERA numbers, and keep approval on for first messages.

How much does an AI agent for a real estate business cost?

On KutrimBots, the Solo plan is $20 a month for up to five agents with the CRM, Bookings, Forms and Chat Widget included, as of November 2026. See the pricing page for the Team and Business plans.

Sources

Keep reading

Use cases 5 min read

CRM lead follow-up on autopilot with AI agents

Most leads are lost to slow or forgotten follow-ups. See how AI agents capture leads, qualify them with BANT and follow up on time, with your approval.