Real estate

Every lead matched to a property and followed up

An agency brings leads from property portals and its website into one CRM, matches buyers to listings, schedules viewings and has listing descriptions drafted in several languages for an agent to approve.

Real-estate agency, 12 agents

Buyers

New lead 2

Couple, two-bedroom flatPortal enquiry · German

Investor, building plotWebsite form

Viewing booked 2

Family, house with a gardenSaturday · two properties

Buyer from abroadVideo viewing

Offer 1

Flat near the centreOwner is considering the offer

Tools we typically use: Custom CRM, Pipedrive or HubSpot, n8n or Make, Claude or GPT models, Next.js, PostgreSQL, Google Workspace or Microsoft 365
Services involved: CRM implementation, AI integrations, Business automation

The build at a glance

4lead sources in one CRM

3languages drafted for each listing

1approval by an agent before a listing is published

1report per owner, prepared every month

The sector

An agency works for two sides at once: owners who want to know what is being done to sell their property, and buyers who ask about several properties at the same time.

Leads come from property portals, the agency’s website, social media and referrals, each in its own format. A listing has to be written, translated and published in several places. Viewings need the agent, the buyer and often the owner on the same day. Whoever answers first and follows up consistently usually gets the client.

Where the time goes

Portal enquiries arrived as emails to individual agents and stayed in their inboxes. Each agent kept their own list of buyers and what they were looking for, so a new listing was matched from memory. Descriptions were written once and translated when somebody had time. Viewings were arranged in long message threads. Owners called to ask whether anything was happening, and the agent put together an answer from memory.

Leads from portals sit in personal inboxes
Matching buyers to new listings depends on one agent’s memory
Listing texts and translations are written by hand, late
Owners have no regular view of the interest in their property

One system around how the work is done

Listing description draft

Write the description for the two-bedroom flat with a sea view, in English, German and Croatian.

Here are three drafts based on the **property record**: floor area, floor, orientation, parking and year of renovation. I have left out the distance to the beach because it is not in the record. Please check and add it.

Property record read Three language versions drafted Waiting for the agent’s approval

01
Leads from portals, the website and social media collected in one CRM, with the property they asked about and their language recorded
02
Buyer requirements stored as fields, so that a new listing is matched to buyers who are looking for it
03
A first reply drafted in the buyer’s language, which the agent approves before it is sent
04
Listing descriptions drafted from the property data in several languages, checked and edited by the agent
05
Viewing scheduling from the agent’s calendar, with confirmations and reminders for the buyer and the owner
06
Follow-up tasks after each viewing, so that no buyer is left without an answer
07
A monthly report for each owner: enquiries, viewings and feedback

How the project runs

Four steps, one process at a time

01
Map the process
We follow a few recent sales from the first enquiry to the contract and list every place where leads arrive.
02
Pilot one process
Lead capture from one portal and the website goes live first, with two agents, including the drafted first reply.
03
Extend
Matching, viewing scheduling and listing drafts follow, then the reports for owners.
04
Care
We adjust the integrations when portals change their formats and refine the matching rules with the agents.

What changes for the team

Leads are shared by the agency, not kept in private inboxes
Agents edit drafted descriptions and replies instead of writing them from scratch
Viewings are confirmed and reminded without message threads
Owners see enquiries, viewings and feedback every month

Every lead is in the CRM with an owner, whichever portal it came from. A new listing brings up the buyers it suits, and the agent decides whom to call. Owners receive a regular report instead of having to ask.

A composite scenario based on typical projects in this sector. It does not describe a named client, and the sample data is illustrative.

The same working day, before and after

What a few ordinary moments look like once the system is in daily use.

BeforeAfter

An enquiry arrives from a portalIt lands in one agent’s inbox and waits until they are back from a viewing.It appears in the CRM with the property attached, and a reply draft is ready for approval.

A new property is listedThe agent tries to remember who was looking for something similar.The CRM lists buyers whose requirements match, and the agent chooses whom to contact.

Writing the listingOne description is written and translated later, if at all.Drafts in three languages are prepared from the property data and edited by the agent.

Arranging a viewingTimes are agreed in a long thread with the buyer and the owner.The buyer picks from free slots, and both sides get a confirmation and a reminder.

An owner asks for newsThe agent recalls the calls and viewings from memory.The monthly report already shows enquiries, viewings and what buyers said.

Questions from this sector

Can you pull leads from the portals we advertise on?

Usually, yes. Some portals offer an interface, others send a structured email that can be read reliably. We check each portal you use during the mapping and tell you which route applies.

Can listings be published to portals from one place?

Where a portal accepts a feed or an interface for listings, yes. Where it does not, the approved text and photos are prepared so that publishing by hand is a matter of pasting.

Are AI-written descriptions accurate?

The draft uses only the data on the property record and says what is missing rather than inventing it. An agent reads and approves every description before it is published, in every language.

Our agents work on commission and guard their contacts. How does a shared CRM fit?

Each lead has an owner, and visibility can follow your rules: an agent sees their own buyers, the office manager sees everything. We set this up to match how you already divide the work.