Real estate agencies receive very different enquiries — sellers, buyers, rental applicants, existing tenants, landlords and maintenance reports. Sending every caller to the same person creates friction. This guide shows how to route each enquiry type correctly, with clear limits on what an AI receptionist should never do, and how every call can reach the right next step.
A real estate office may receive several calls within a short period. Caller one says, "I'm thinking of selling my property." Caller two says, "I'm interested in the property on Smith Street." Caller three says, "I'd like to apply for a rental." Caller four says, "I'm an existing tenant and need to report a maintenance issue." Caller five says, "I'm a landlord looking for property management."
All five called the same agency. But they should not all enter the same workflow.
The first useful job of reception is often not to solve the enquiry. It is to work out who should handle it.
Potential vendor or homeowner.
Interested in buying or inspecting property.
Looking for available rental property.
Current tenancy enquiry.
Owner seeking management services.
Current property-management client.
Existing tenant or landlord reporting an issue.
Office hours, location, staff member, administrative question.
Different enquiries may need different teams, different CRM pipelines, different questions, different privacy handling and different escalation processes.
An agency may have:
Sending every caller to "the next available agent" may create unnecessary handoffs. A clearer intake process can first determine:
What is this enquiry actually about?
The preferred hybrid model works like this:
Agency phone rings normally → reception / agent answers? YES → human conversation. NO → configured no-answer forwarding → AI receptionist → enquiry category identified → relevant details captured → CRM / appropriate system → correct person notified → human follow-up.
Your office phone still rings first. If you can't answer, the AI can take over.
Forwarding delay depends on the carrier, VoIP, PBX, phone system and configuration. Do not promise exactly three rings universally. For the broader explanation, see our guide on what an AI receptionist is and how it works and the AI receptionist service.
Use something like:
"How can I help?"
Then identify broad intent. Potential routing:
One phone number can lead to several organised pathways.
Potentially capture:
Example:
Seller enquiry → sales / appraisal pipeline → agent notification → human follow-up.
Do not have the AI independently estimate:
This is mandatory. The AI should not say:
Unless the system is simply relaying an explicitly approved existing published figure for a specific legitimate context, valuation and pricing discussions belong with appropriate professionals. The AI can instead:
Capture the appraisal request → route it to the agent.
AI can book the appraisal conversation. It should not pretend the appraisal has already happened.
Potential information:
Possible pathway:
Buyer → specific property? YES → relevant listing agent / team. NO → buyer enquiry / general sales team.
Do not collect unnecessary financial information during generic reception.
This is critical. If the customer asks:
"Is 10 Smith Street still available?"
The AI should answer only if current approved listing information is genuinely available from a reliable connected source. Otherwise: capture / transfer / notify the listing agent. Do not invent:
If the source isn't reliable and current, don't make up the property information.
Where current inspection schedules are genuinely connected, AI may potentially:
But: do not invent inspection availability. Do not promise private inspection availability without agent confirmation.
Separate:
from:
These require completely different handling.
Prospective tenant questions might involve:
Existing tenant questions might involve:
Do not mix these into one generic rental workflow.
Potential intake:
Do not have a generic AI receptionist make:
Application assessment belongs with the agency's approved process and relevant people / systems.
This is an important privacy boundary. A general receptionist does not need to collect:
simply to answer an initial rental enquiry. Instead:
Initial enquiry → approved application process / secure system.
The OAIC says organisations covered by the Privacy Act should only collect personal information reasonably necessary for their functions or activities, and its tenancy guidance specifically addresses real-estate agents' handling of renter information.
Reception should collect enough to route the enquiry — not recreate the tenancy application.
Potential categories:
The AI may identify the category, capture appropriate information and route to property management. It should not interpret tenancy law.
Do not allow the AI to independently answer questions such as:
Tenancy laws vary by Australian state and territory and depend on the facts. The AI can:
Do not position AI as a lawyer or tenancy authority.
Example:
Existing tenant → maintenance → capture tenant name, property / address as appropriate, contact, description in tenant's own words, photos / link only through approved process where relevant, caller-stated urgency → property management workflow → human / approved maintenance process.
Do not have AI determine legal responsibility for the repair.
Tenant says:
"There's water coming through the ceiling."
Record:
Tenant reports water entering through ceiling.
Do not automatically claim:
The agency or appropriate contractor determines what has actually occurred.
Tenancy rules and definitions can vary by state and territory. The AI should not invent a national list and independently make legal determinations about:
Instead:
Tenant describes issue → AI captures caller-reported details → agency's approved escalation rules → property manager / authorised process.
Where immediate threats to life or genuine emergencies exist, the agency's approved process may direct people to relevant emergency services. Do not give DIY repair instructions.
Do not let a general AI receptionist tell a tenant, contractor or landlord that repairs have been approved unless the authority genuinely exists in the connected system / process. It should not invent:
Capturing a maintenance request is not the same as authorising work.
Potential categories:
A new landlord enquiry may potentially capture:
New landlord enquiry → property management BDM / relevant team.
Do not provide property-management legal advice.
Appraisal → sales agent.
Property management enquiry → property management / BDM.
They are different commercial opportunities. A CRM can route them separately. See CRM and lead management.
Existing owner enquiries may involve:
Do not funnel them into a new-business sales workflow. This is why existing / new status can be useful during initial routing.
Examples:
AI can:
Then:
Human takes over.
Do not configure the AI to argue, defend the agency, determine fault or provide legal conclusions.
If someone asks:
"Can I speak with Sarah?"
the system should:
according to the agency's process. Do not make the caller answer ten qualifying questions first.
Routing should remove friction — not create another gatekeeper.
Keep data minimal and purpose-based. Potential general fields:
Different pathways may require different fields. Do not collect information simply because CRM supports it.
Real-estate businesses can handle substantial personal information. Examples may include:
The OAIC says a covered organisation should collect only personal information reasonably necessary for its functions or activities, and its tenancy guidance specifically cautions against collecting unnecessary renter information. Do not turn a generic marketing CRM into an uncontrolled repository for detailed tenancy-application documents.
Keep this concise and current.
From 1 July 2026, certain services provided by real-estate professionals became subject to Australia's expanded AML/CTF regime. Current OAIC guidance also addresses privacy obligations associated with these reforms.
Therefore, JMG Nitro should not position a simple reception / CRM workflow as a substitute for:
If an enquiry progresses to a regulated transaction / process, move it into the agency's approved compliant systems. Do not claim JMG Nitro certifies AML/CTF compliance. Use current OAIC / AUSTRAC guidance where discussing this topic.
This distinction is mandatory. Real-estate agencies may already use specialist platforms for:
A general CRM should not automatically be positioned as a replacement. It may instead support:
Integrate around specialist systems where useful — don't duplicate the whole agency operation unnecessarily.
Example:
New enquiry → seller → appraisal pipeline. Buyer → buyer / listing agent. Landlord → property management lead. Rental applicant → leasing process. Existing tenant → property management service.
This is where CRM categorisation becomes useful.
Caller:
"I'm considering selling my house."
AI collects name, phone, suburb / property where appropriate and preferred contact time → sales agent → appraisal conversation. Do not provide valuation.
Caller:
"I'm interested in the Smith Street property."
AI identifies listing, captures contact and records question. If current approved listing information exists, provide only that information. Otherwise → listing agent notification / transfer. Do not hallucinate property details.
Caller:
"When can I inspect this rental?"
If current inspection times are connected and approved, AI may provide those. Otherwise → capture interest → leasing team. Do not invent inspection times, property availability or application outcome.
Tenant:
"The air conditioner isn't working."
AI identifies existing tenant, captures property, records tenant's description and follows approved maintenance routing → property manager / approved process. Do not diagnose the HVAC fault, decide tenant / landlord responsibility or approve contractor expenditure.
A real-estate website may use:
Then:
Form → CRM / appropriate system → correct team.
See website design and our guide on how website enquiries can trigger SMS and email follow-up.
Human / AI → enquiry category → correct team.
Form → enquiry category → correct team.
Then:
CRM / specialist system → human follow-up.
This creates a clearer customer journey.
Suitable administrative workflows may include:
See practical business automation and SMS & email follow-up. Do not fabricate claims about higher attendance or conversion.
Possible options:
AI may handle after-hours:
Existing tenant matters may require different escalation.
AI being available after hours does not mean an agent, property manager or contractor is immediately available.
AI should never imply:
unless supported by current connected information.
For real estate, slightly extend the framework:
Google, website, Local SEO.
Reception, agent, AI backup.
Buyer, seller, rental, landlord, tenant.
Agent / property manager, CRM, specialist system.
See Local SEO for Australian businesses and the Local Growth System.
A local agency's Maps visibility may vary across its geographic market. A geo-grid heatmap can visualise local visibility from different search locations for selected queries.
Do not claim:
Use the heatmap as diagnostic visibility information.
Automation can:
Automate the routing — not the property judgement.
Agency phone rings normally → human answers if available → if unanswered: AI receptionist backup → identify seller / buyer / rental / tenant / landlord / other → route → CRM / specialist system → human follow-up.
Visitor selects enquiry category → form → same routing framework.
This should be the strongest practical takeaway.
Voicemail may still suit an agency where:
Do not imply every agency needs AI.
Potential indicators:
These are indicators — not guaranteed ROI.
Review whether generic enquiry workflows are collecting unnecessary personal information.
Do not call this an AML compliance audit, privacy certification or real-estate compliance audit. JMG Nitro is reviewing the digital / enquiry process. Describe privacy / compliance findings as areas for the agency to review with its appropriate advisers / process owners. This may later support JMG Nitro's planned Free Local Growth Audit without changing the sitewide CTA yet.
For related reading, see our guide on what an AI receptionist is and how it works, our comparison of AI receptionist vs voicemail, our guide on what an AI receptionist should ask a new enquiry, our guide on what a CRM is and whether a small business needs one, and our guide on which small business tasks to automate. To see how these systems fit a real estate agency specifically, visit our real estate industry page.
An agency cannot always answer every call — but every enquiry can still reach the right person. See how JMG Nitro configures call handling, AI backup, CRM and follow-up around a real estate agency's real workflow, without the AI ever valuing property or giving tenancy advice.
Other practical guides connected to this topic.
A plain-English explanation of how an AI receptionist captures enquiries that can flow into a real estate agency's CRM.
Read guideAI Reception & Call HandlingA fair comparison of AI receptionists and voicemail — including where voicemail may be perfectly adequate for an agency.
Read guideAI Reception & Call HandlingHow to collect useful initial information from a real estate caller without turning the call into an interrogation — and without collecting unnecessary tenancy-application data.
Read guideCRM, Automation & Follow-UpWhere agency enquiries live once they arrive — and how a CRM complements rather than replaces specialist real-estate software.
Read guideBook a free demo to walk through how JMG Nitro's systems could fit around your existing business process.
0485 099 000
Back to Resources