Skip to content
StriqTech Logo
Back to blog
Automation9 min read

How to automate a real estate agency with n8n: leads, viewings, and CRM (2026)

How to automate a real estate agency with n8n to unify portal leads, integrate Tokko and Zonaprop, and cut time-to-first-response. 2026 guide.

Automating a real estate agency with n8n lets you integrate Tokko, Zonaprop, and WhatsApp into a single flow, qualify leads before they reach a salesperson, and cut time-to-first-response.

TL;DR

Why n8n is the right platform for Argentine real estate agencies in 2026

The real pain of any mid-sized real estate agency isn't generating leads: it's attending to them in time. The portals (Zonaprop, Argenprop, MercadoLibre, TokkoBroker feed) push contacts via email, WhatsApp, forms, and each one lands in a different channel. The human salesperson takes on average 4h to respond. Meanwhile the customer has already talked to 3 other agencies.

n8n solves this because:

  1. It unifies channels. Scrape + webhook + API in a single workflow.
  2. It runs 24/7. The lead from Sunday 23:30 gets a response before 8am Monday.
  3. It integrates with the regional stack. Tokko, Zonaprop (via email parsing or scraping), MercadoLibre API, WhatsApp.
  4. Self-hosted. Your customer data stays on a server you control, which helps you comply with Law 25.326.

The 5 most profitable processes to automate in real estate

Prioritized by revenue impact, not by technical ease.

1. Portal → unified CRM sync

Zonaprop, Argenprop, and MercadoLibre each send their leads through their own channel. n8n centralizes them:

  1. Zonaprop webhook → parser → Tokko.
  2. Argenprop email to alias inbox → IMAP node → parser → Tokko.
  3. MercadoLibre API → polling every 5 min → Tokko.
  4. Own web form → webhook → Tokko.

All leads enter with the same structure and tagged source. No more duplicates across channels, no more "did Juan already call this lead?".

2. Capture via WhatsApp bot with per-property response

Every published listing carries a "Inquire via WhatsApp" button with the property code embedded. The customer clicks → sends a pre-formatted WA → bot identifies the property via Tokko API → responds with price, expenses, age, 3 photos, and a 360 tour link.

Then: "Want to schedule a viewing this week?". If they say yes, it moves to flow 4.

3. Automatic lead qualification

Not all leads are equal. n8n runs scoring on each new lead:

  1. Declared budget vs property range: if they said 50k USD and the property costs 180k, it drops to nurturing.
  2. Searched area vs inventory: direct match, hot lead.
  3. Required financing vs possible: discards incompatible leads.
  4. Previous interaction history: if they already inquired about 3 different properties, it's prioritized.

Output: scoring 0-100. Only 70+ scores pass to the human salesperson. The rest go to automatic nurturing (WhatsApp drip campaign).

4. Viewing scheduling with Google Calendar

Customer confirms a viewing → n8n queries the salesperson's availability in Google Calendar → proposes 3 slots → customer picks → an event is created with location link, WhatsApp confirmation sent, reminder 24h before, reminder 2h before.

If there are 2h left and the customer hasn't confirmed, automatic call or message to reconfirm. No-show reduction: from 25% to 8%.

5. Post-viewing follow-up and nurturing

Key automation 90% of agencies don't do. After the viewing:

  1. 2 hours later: WhatsApp with souvenir photos + "What did you think?".
  2. Day 3: recommended similar properties.
  3. Day 7: if no interest, quick objection survey.
  4. Day 30: commercial check-in.
  5. Day 90: low-frequency nurturing with area news.

Salespeople don't do this manually. n8n does, and that's where the extra 30% of closings come from.

n8n workflow step by step: Zonaprop → PostgreSQL → Tokko → WhatsApp bot

An example flow, node by node.

  1. Trigger: IMAP email node connected to the inbox receiving Zonaprop notifications.
  2. Code node (avoid internal fetch, only parsing): extract name, phone, email, property_id, message fields from the email HTML.
  3. Set node: normalize phone to E.164 format (+54 9 11...) and generate a SHA256 hash of phone+email as the lead fingerprint.
  4. PostgreSQL node: query whether that fingerprint exists in the leads table. If it does, route to "update lead". If not, route to "create lead".
  5. HTTP Request node to Tokko API (POST /api/v1/contacts/) to create the contact with source="zonaprop".
  6. HTTP Request node to Tokko to fetch data of the property they inquired about.
  7. HTTP Request node to Evolution API (/message/sendText/{instance}) with template: "Hi , this is Propiedades XYZ. I saw you inquired about . Want full info or to schedule a viewing?".
  8. PostgreSQL node: store the first-response timestamp for the TTR metric.
  9. Wait node: if there's no response within 2h, send a reminder.
  10. Slack node or WhatsApp to the salesperson: notify the human there's a hot lead waiting.

Build time depends on the systems involved and is set in the proposal.

The categories and tools we've seen work in real implementations.

Real estate CRM

  1. Tokko Broker — AR leader, complete REST API, multi-portal publishing.
  2. HubSpot — more complete in nurturing and reporting, integrates with everything.
  3. Infocasas CRM — regional alternative with a good API.

Publishing portals

  1. Zonaprop — highest volume in AR, lead via email + webhook if you have the pro plan.
  2. Argenprop — second in volume, good lead quality.
  3. MercadoLibre Inmuebles — official API, reliable webhooks.
  4. Remax / Century21 internal — if you operate within the network.

Communication

  1. WhatsApp via Evolution API (MVP) or Cloud API (production) — main channel.
  2. Chatwoot — multi-agent dashboard for salespeople.
  3. Twilio SendGrid — transactional and nurturing email.

Calendar and digital signature

  1. Google Calendar — viewing scheduling.
  2. Cal.com self-hosted — alternative with public booking.
  3. DocuSign or HelloSign — digital signature for reservations and deeds.

Analytics

  1. Looker Studio — dashboard for TTR, conversion by source, pipeline.
  2. PostgreSQL on your own server — central database.

Hypothetical scenario: LATAM real estate agency with Tokko + n8n, 2x qualified leads

Suppose a real estate agency in Palermo, team of 6 salespeople, inventory of 180 active properties, commission revenue USD 28k/month.

Pains before:

What would be implemented:

  1. Deploy n8n + Evolution API + PostgreSQL on Hostinger KVM 4 (USD 15/month).
  2. 5 workflows: portal sync → Tokko, WhatsApp bot with Tokko API, automatic scoring, scheduling with Google Calendar, post-viewing drip.
  3. Looker Studio dashboard with TTR, conversion, pipeline by salesperson.

90-day targets under these assumptions (assumptions, not measured results):

How much it costs to automate your real estate agency with n8n

It depends on how many portals, systems, and salespeople are in the loop, and on how much has to be integrated with Tokko or built custom (scoring, reporting, multiple branches). At StriqTech we work in three steps: first, a no-commitment conversation to understand how leads come in and get handled today; then a proposal with scope, deliverables, dependencies, acceptance criteria, stages, and timelines; and finally we build, test with the people in charge, and agree on go-live and maintenance. The proposal spells out which maintenance, hosting, or third-party services are included.

5 common mistakes when automating real estate

  1. Duplicating leads across portals due to missing fingerprint. The same customer comes in via Zonaprop and MercadoLibre, ends up with 2 contacts in Tokko and 2 different salespeople calling them. Fix: hash of phone+email before creating in CRM.
  2. Not qualifying the lead before passing it to a human. Salespeople lose 40% of their time with leads that have no budget. Fix: automatic scoring with LLM + declared budget + area.
  3. Manual follow-up after the viewing. The salesperson promises to call and doesn't. Fix: automatic drip on day 0, 3, 7, 30, 90.
  4. Not measuring TTR. If you don't measure time-to-first-response, you can't improve it. Fix: Looker Studio dashboard with TTR by channel and by salesperson.
  5. Not integrating Google Calendar in scheduling. Customer asks for a viewing, salesperson promises to confirm, never does. Fix: automatic slot based on the salesperson's real availability.

CTA

If your real estate agency loses leads because of slow response times or because salespeople load Tokko by hand, tell us which process you want to fix. After a first conversation we'll propose the next step, with scope, stages, and timelines in writing. Write to us at +54 9 11 5499-7296.

Frequently asked questions

Is it worth automating if I'm a one-person real estate agency?

Yes, especially as a one-person team. The bottleneck isn't lack of leads but the time to attend to them. n8n running 24/7 qualifies leads while you sleep and in the morning you have a prioritized list.

Can I connect Tokko with n8n?

Tokko exposes a REST API with endpoints for properties, contacts, and activities. From n8n you use HTTP Request with the Tokko token to create leads, update statuses, and sync properties with portals.

How do I avoid duplicating leads between Zonaprop, Argenprop, and MercadoLibre?

With a lead fingerprint: hash of normalized phone + email. n8n checks if it exists in PostgreSQL before creating in Tokko. If it exists, it appends the source and message to the lead; if not, it creates a new one.

Can the WhatsApp bot answer questions about properties?

Yes. GPT-5 or Claude receives the message, identifies the property by code or description, queries Tokko via API, and builds a response with price, expenses, photos, and a 360 tour link. Then it offers to schedule a viewing.

How long does implementation take?

It depends on how many portals and systems need to be integrated: a narrow setup (portals, WhatsApp bot, viewing scheduling, post-viewing follow-up) doesn't take the same time as a deep Tokko integration with reporting. Timelines are set in the proposal, by stage, separating StriqTech's work from what depends on the agency and on third parties.

Does it comply with the personal data law (Law 25.326)?

Yes if implemented properly: lead data stays on a server under your control, with explicit opt-in on forms and a published privacy policy. Where it is hosted and how automatic retention and deletion are configured is defined in the proposal.

Can I measure how long it takes me to respond to each lead?

n8n stores the timestamp of receipt and of the first response in PostgreSQL. In Looker Studio you build a dashboard with TTR (time to response) by channel and by salesperson. It's the metric that moves the needle.

Want to apply this in your company?

Tell us which process you want to sort out and which systems you use today. We review your inquiry and coordinate the next step.

Let's talk on WhatsApp

You may also like