Respond to Property Enquiries 24/7: AI Assistant Guide
Published
As of 17 September 2026, an AI assistant on WhatsApp lets an estate agency respond to property enquiries 24/7 — answering questions about parking, pets, EPC ratings and lease length in under 3 seconds, booking viewings straight into the diary, and escalating negotiations or complaints to a named human. Nothing sits unanswered overnight, including weekends and bank holidays.
An AI assistant for property enquiries is a software agent that reads incoming messages on WhatsApp, Telegram, Messenger and Instagram, answers listing and process questions from your own agency data, books viewings into your calendar, and passes anything sensitive to a human. It runs every hour of every day, including weekends.
Key takeaways
- An AI assistant answers routine property enquiries in under 3 seconds, at 10pm on a Sunday as readily as at 10am on a Tuesday.
- It books viewings directly into the diary, sends confirmations, and chases reminders — the three steps where most agencies lose momentum.
- It qualifies leads on budget, buying position, chain status and mortgage in principle before a human ever picks up the phone.
- Human escalation is mandatory, not an optional extra: negotiations, complaints and anything needing judgement go to a person.
- Setup runs on sector templates and takes days, not months — you map your top 20 enquiries and your booking rules, then switch it on.
Why can't estate agents respond to property enquiries 24/7?
Photo by https://kaboompics.com/ on Pexels
Estate agents cannot respond to property enquiries 24/7 because the front desk is staffed by people who are also doing viewings, valuations, key handovers and paperwork. A two-person agency covering 40 listings has roughly 80 working hours a week between them; enquiries arrive across 168 hours. The gap is where instructions are lost.
The 10pm enquiries that go to whoever answers first
Property enquiries peak outside office hours. A buyer scrolling Rightmove at 10pm sends three messages — one to you, one to the agency down the road, one to a competitor two towns over. The first reply usually wins the viewing. If your office opens at 9am, you are replying 11 hours late, and by then the buyer has booked with whoever answered at 10:04pm.
This is the single most fixable leak in an estate agency. The enquiry is already warm, already qualified by the listing itself, and already interested enough to type a message. It just needs an answer.
Viewing requests that land while you're mid-appointment
A viewing request that arrives while you are standing in a client's kitchen does not wait. It sits in a notification tray until you are back in the car, then until you are back at the desk, then until after the next appointment. Three viewings requested in a morning can become three viewings booked for tomorrow afternoon — or three viewings booked with someone else.
The same problem hits valuations and landlord instructions. A landlord who wants a rental appraisal at 8pm on a Friday is not going to try again on Monday.
What a missed enquiry actually costs an estate agency
A missed enquiry costs an estate agency the instruction, not just the conversation. If your average sale fee is £2,500 and you lose two instructions a month to slow replies, that is £60,000 a year in fees that went to another agency — before counting the vendor who never got a valuation booked.
The arithmetic is unforgiving because the loss is invisible. Nobody complains about a message that was never answered. You simply see a quieter pipeline and assume the market is slow.
What can an AI assistant handle for a property business?
An AI assistant for a property business handles the first 80% of every enquiry — the repetitive, factual, bookable part — and hands the remaining 20% to a human. That split covers listing questions, viewing bookings, lead qualification, maintenance reports and reminders, on every channel you connect.
Instantly answering the questions every enquirer asks: parking, pets, EPC, lease length, school catchment
Every enquirer asks the same eight or nine things: parking, pets, EPC rating, lease length, service charge, council tax band, school catchment, garden aspect and whether the boiler is new. An AI assistant answers all of them from your listing data in under 3 seconds, at any hour, without a member of staff typing a word.
The assistant only answers from what you have given it. If the EPC rating is not in the listing, it says so and offers to check with the negotiator — it does not guess. That restraint is what makes the answers trustworthy enough to publish.
Booking viewings straight into the diary and sending confirmations
An AI assistant books viewings by reading your availability, offering two or three slots, booking the chosen one into the calendar, and sending a written confirmation with the address and the negotiator's name. There is no round trip through the office and no "I'll come back to you with times".
The sequence runs in four moves: pull free slots from the calendar, offer two or three, write the chosen slot back as a confirmed appointment, then send the confirmation message. Each one is a single step, so a request that arrives at 22:40 on a Sunday is a booked viewing before midnight — not a note for Monday morning.
Confirmations are not optional. A written confirmation carries the property address, the date and time, the negotiator's name and a cancellation link, which is what stops the viewer turning up at the wrong address at the wrong hour. If the negotiator's diary changes, the assistant re-offers the slot rather than leaving a silent gap.
For agencies that want this handled end to end, the pattern is set out in this guide to booking property viewings automatically with AI assistants.
Qualifying leads: budget, buying position, chain status, mortgage in principle
An AI assistant qualifies leads by asking budget, buying position (first-time buyer, cash, chain), chain status and whether a mortgage in principle is in place — then tagging the enquiry so your negotiator knows who to call first. A cash buyer with no chain and a mortgage in principle is a different phone call from a browser with no budget.
The four questions are deliberately fixed and asked in the same order:
- Budget — a figure, and whether it includes fees and stamp duty, or is the ceiling on the asking price alone.
- Buying position — first-time buyer, cash purchaser, or already in a chain; cash buyers skip the lender entirely, which changes the timeline.
- Chain status — nothing to sell, a property under offer, or a property not yet listed; the third is the slowest and the most likely to fall through.
- Mortgage in principle — in place and with which lender, in progress, or not started.
The answers become a tag on the enquiry, so the negotiator's call list is ordered by readiness rather than by arrival time. The full question set and scoring approach is covered in this breakdown of WhatsApp lead qualification for real estate.
Taking maintenance reports and landlord instructions out of hours
Tenants report leaks, broken boilers and damp at the worst possible times. The assistant logs the report, asks for photos, captures the address and urgency, and routes it to the right person or contractor. Landlords asking for a rental appraisal or a management quote get the same treatment: details captured, appointment offered, file created.
Sending reminders that cut no-shows for viewings and valuations
Reminders are the cheapest no-show reduction available. The assistant sends a confirmation when the viewing is booked, a reminder 24 hours before, and a short nudge on the morning. If the viewer cancels, the slot reopens automatically and the next enquirer on the list gets offered it.
How does an AI assistant customise to your agency's services?
Photo by Markus Winkler on Unsplash
An AI assistant customises to an agency's services by starting from a sector template and then learning your listings, your fees, your process and your tone. Nothing is hard-coded to a generic script — the assistant answers as your agency, about your stock, using your rules.
Built from sector templates, not generic scripts
The assistant is built from sector templates, which means the first version already understands sales and lettings vocabulary, viewing logistics and the standard compliance questions. You are not teaching it what an EPC is. You are teaching it what your EPC ratings are.
Your listings, your fees, your process — the assistant learns them
You load your listings, your fee structure, your viewing rules (do you accompany? do you use keys? how much notice?) and your office hours. The assistant then answers within those boundaries. If your fee is 1.2% plus VAT, it says 1.2% plus VAT. If you do not discuss fees before a valuation, it does not either.
Works across WhatsApp, Telegram, Messenger and Instagram
The assistant runs on WhatsApp, Telegram, Messenger and Instagram, so an enquirer can start on Instagram, continue on WhatsApp, and never repeat themselves. WhatsApp compliance has been a live requirement since January 2026, and tenant data isolation is built in — one agency's data is never visible to another.
For a closer look at what this looks like in practice for sales and lettings teams, see AI assistant for estate agents: book more viewings.
When does the assistant hand over to a human?
The assistant hands over to a human whenever an enquiry involves negotiation, complaint, legal risk or judgement. Human escalation is a mandatory feature of the system, not an optional add-on — the assistant is designed to know the edge of its own competence and stop there.
Negotiations, complaints and anything that needs judgement
Offers, counter-offers, fee negotiations, boundary disputes, service charge arguments, anything involving a solicitor, and any message containing the words "complaint" or "unacceptable" go straight to a named person. The assistant does not improvise on price or on liability.
Why is human escalation mandatory rather than an add-on?
Human escalation is mandatory because an assistant that never hands over is a liability. An offer is a legally significant communication, and a complaint carries regulatory weight — in both cases the agency, not the software, is accountable to the client and to its redress scheme. An assistant that answers an offer with a plausible-sounding figure has given advice nobody authorised.
The handoff itself is a defined package, not a forwarded message. When a message is escalated, the human receives the full thread, the enquirer's details and a short summary of what has already been answered — so the caller never has to start again. Outside handoff hours, the assistant tells the enquirer when a person will reply and keeps the thread open.
Escalation triggers are set by you, and they can be as specific as your business needs: a fee threshold, a named competitor, a solicitor's name, or the words "complaint" or "unacceptable". The assistant does not decide where its competence ends — your agency does.
Before and after: the front desk with and without an AI assistant
Photo by Wafiq Raza on Unsplash
The table below compares a typical two-person estate agency front desk before and after adding an AI assistant, across missed enquiries, no-show rate, response time and after-hours bookings. Figures describe the operating pattern, not a guaranteed result for every agency.
| Measure | Without an AI assistant | With an AI assistant |
|---|---|---|
| Missed enquiries | Enquiries arriving after 6pm wait until the next working day | Every enquiry answered, including weekends and bank holidays |
| Average response time | Hours, often the next morning | Under 3 seconds |
| No-show rate for viewings | Reduced only by manual reminder calls | Cut by automated 24-hour and same-day reminders |
| After-hours bookings | Zero — viewings booked only in office hours | Viewings booked at 10pm, on Sundays and on bank holidays |
| Lead qualification | Done on the phone, when someone is free | Done in the first message exchange, before a human is involved |
How to go live in a few days
Going live takes days, not months, because the assistant starts from a sector template rather than a blank page. The work on your side is mapping enquiries and rules, not building software.
Checklist: map your top 20 enquiries and your booking rules
Write down the 20 questions your team answers most often, with the exact answer you want given. Then write your booking rules: viewing slots, notice period, who accompanies, and what happens if a slot is taken. Two hours of writing here saves weeks of correction later.
Choose your channels (WhatsApp, Telegram, Messenger, Instagram)
Connect the channels your enquirers actually use. Most agencies start with WhatsApp and Instagram, then add Messenger and Telegram if their audience is there. The same assistant and the same knowledge base serve all four.
Set your escalation triggers and handoff hours
Decide what must always reach a person: offers, complaints, legal questions, and anything above a fee threshold you set. Then set handoff hours — the window in which a human replies — and make sure the assistant states those hours honestly.
Test with real enquiries, then switch it on
Run 20 real enquiries through the assistant, including three awkward ones, and check every answer against your listings. When the answers are right, switch it on. The full picture of what an AI front desk does for a clinic — same mechanics, different sector — is in this piece on after-hours answering for medical clinics.
María R., a patient at Clínica Dental Sol in Barcelona, described the change in her own words: their WhatsApp used to go unanswered after 6pm, and now every patient gets a reply in seconds and books themselves. The same shift applies to a viewing request sent at 10pm.
Frequently asked questions
Can the assistant book viewings directly into my calendar?
Yes. The assistant reads your availability, offers real slots, books the chosen one into your calendar, and sends a written confirmation. If the viewer cancels, the slot reopens and is offered to the next enquirer automatically.
What happens if it can't answer a question?
The assistant says it will check and escalates the message to a named person with the full thread attached. It does not guess at facts it has not been given. The enquirer is told when a human will reply.
Does it work on Instagram and Facebook Messenger as well as WhatsApp?
Yes. The assistant runs on WhatsApp, Telegram, Messenger and Instagram from one knowledge base, so an enquirer can move between channels without repeating information. WhatsApp compliance requirements in force since January 2026 are handled within the platform.
Is it safe for tenant and client data?
Tenant data isolation is built in, so one agency's data is never visible to another. Escalation to a human is a mandatory feature, which keeps judgement calls — and accountability — with your team.
How long does it take to set up?
Setup takes days, not months. Most of the time goes into mapping your top 20 enquiries, your booking rules and your escalation triggers; the assistant itself is pre-trained for SMEs and built from sector templates.
Ready to let your business answer for itself? Start with an AI assistant that answers and books.
Last updated 17 September 2026