How to answer buyer questions on price and availability after hours

This page may contain affiliate links from which Blur Studio earns commissions.

Property buyers don't keep office hours when they're actually interested. A listing catches their attention in the evening, they call or message right then, and if nobody answers, a lot of them move to the next listing, not back to yours tomorrow. The question isn't whether to respond fast, it's how to do it without staffing someone around the clock.

Your options

• A live chat team: real people answering in real time, which handles nuance well but means paying for coverage at hours when call volume is genuinely low most nights.
• An AI voice agent: answers the phone instantly every time, follows a script covering the questions that come up constantly (price, availability, basic specs, how to book a viewing), and should hand off anything involving negotiation, legal advice, or a question only a licensed agent should answer.

Where an AI voice agent fits

ElevenLabs' agent platform builds a phone agent from a written script, a voice, and a knowledge base of your actual listing details, then connects it to a number buyers can call any time. Pro is $99 a month with roughly 1,238 minutes of agent time included, Scale is $299 with about 3,738 minutes, and Business is $990 with about 12,375 minutes; usage past your included minutes runs close to $0.08 a minute, with the language model billed separately. See current ElevenLabs plans before picking a tier, since the right size depends on your actual after-hours call volume.

Load your live listings, price ranges and availability into the agent's knowledge base (documents up to 20MB each, or straight URLs) so it answers from your actual current data rather than a static script that goes stale the moment a listing sells or a price changes. For Dubai brokerages, see how an AI voice agent answers portal calls after hours. The question-by-question setup is in answering property questions with an AI agent.

Why speed to lead matters more than it sounds

Real estate lead response time has been studied enough that the pattern is well established: the odds of actually connecting with and converting a lead drop sharply the longer the response takes, and most of that drop happens in the first hour, not the first day. A buyer who calls at 9pm about a listing and gets an answer right then is still in the mindset that made them pick up the phone in the first place, actively comparing this property against others they're considering. By the next morning, that same buyer has often already moved on, either to another listing or to whichever agent did answer. This is the actual mechanism behind "the first agent to answer gets the lead," and it's why an instant, always-on answer beats a better answer that arrives twelve hours later. For Dubai, and why the popular five-minute figure is shaky, read speed to lead for Dubai brokerages.

What happens after the call

Answering the question is only half the job. A buyer who gets price and availability information at 9pm and hears nothing else until business hours has still lost momentum. Build the agent's workflow to close that gap too: after answering the immediate question, have it offer to book a viewing slot directly if your calendar allows it, or at minimum collect the buyer's contact details and send an immediate text confirmation that someone will follow up first thing. A follow-up text sent the moment the call ends, even a simple "thanks for calling about [listing], [agent name] will reach out by [time] tomorrow," keeps the lead warm in a way that a transcript sitting unread until morning doesn't. This is a small addition on top of the core agent setup, but it's the difference between capturing a lead and just answering a question. Tools for tracking those follow-ups are compared in real estate lead management software.

Step by step

1. List the questions buyers actually ask after hours: price, is it still available, square footage, when can I view it.
2. Write clear answers and load your current listing sheet into the agent's knowledge base.
3. Decide explicitly what the agent should never attempt: negotiating, giving investment advice, or answering anything that requires a licensed broker's judgment.
4. Test the agent yourself at night with real and tricky questions before buyers do.
5. Review call transcripts weekly and update the knowledge base the moment a listing's price or availability changes.

Practical tips

• Disclose that it's an AI, clearly, at the start of the call. ElevenLabs' voices are convincingly close to human, and in many places disclosing that a caller is talking to an automated system is a legal requirement, not just good practice. Build the disclosure into the agent's opening line rather than leaving it implicit.
• Stale listing data is worse than no answer. A confident wrong price or "still available" on a property that sold yesterday damages trust more than the agent saying it will check and call back. Keep the knowledge base synced with your actual current listings, not a snapshot from onboarding.
• Treat latency as part of the buyer experience. ElevenLabs reports a typical response time around 680 milliseconds and up to about 1,560 milliseconds in slower cases. That's close to a real pause in conversation, not instant, so script the agent to acknowledge a question briefly while it "thinks" rather than leaving dead air.

The real cost comparison

A live chat team covering evenings and weekends usually means either paying staff an overnight or after-hours differential, or outsourcing to a chat-staffing service billed per conversation or per hour of coverage; either way, you're paying for availability whether or not a buyer actually calls in a given hour. An AI agent on a flat monthly plan costs the same whether it fields one call or fifty that night, which changes the economics specifically for after-hours coverage, where call volume is real but unpredictable and often low in any single hour. The crossover isn't about total call volume so much as about paying for coverage versus paying for usage: if your after-hours volume is lumpy and unpredictable, a flat-fee agent is a better match for that shape of demand than staffing a person for hours that are mostly quiet.

A mistake worth avoiding

The most common failure mode isn't the agent getting a fact wrong, it's the agent being asked something slightly outside its script and improvising an answer that sounds confident instead of saying it doesn't know. A buyer asking "is the price negotiable" or "why has it been on the market so long" is asking something that needs human judgment dressed up as a factual question. Write explicit rules for the questions you know will come up this way, not just the ones that look factual on the surface, and have the agent default to "let me connect you with the agent handling this listing" whenever it's genuinely unsure, rather than guessing to seem helpful.

What I'd tell an agent or brokerage weighing this

The honest case for this is speed, not replacing a human agent's judgment. A buyer who gets an immediate, accurate answer about price and availability at 9pm is a warmer lead than one who waits until morning, and that gap compounds across every listing you run. What I'd be careful about is letting the AI agent do anything that looks like negotiation or advice, since that's exactly the kind of interaction a buyer, and a regulator, expects a licensed person to handle. Keep the automation strictly to information retrieval and booking, and route everything else to a real person without the agent trying to be helpful beyond its actual job.

It's also worth being honest with your team about what this changes and what it doesn't. The agent doesn't replace the relationship-building part of closing a sale, the part where a buyer decides they trust the person selling them a property. What it does is make sure that relationship gets a chance to start at all, by not losing the buyer to silence before a human ever gets on the phone. Frame it to your team that way rather than as a replacement for their work, since an agent that feels threatening to the sales team tends to get quietly sabotaged (undersold, under-maintained, or routed around) rather than actually adopted.

How I'd approach this

The setup above works well built directly on ElevenLabs' own dashboard. Where I usually get involved is connecting the agent to a live listings feed or CRM, so it's always answering from your current inventory rather than a document someone forgot to update, and logging every after-hours enquiry straight into your pipeline instead of leaving it in a call transcript nobody checks until the lead's gone cold. The follow-up text workflow described above is usually the first thing worth adding once the core agent is live, since it's a small build with an outsized effect on whether a late-night lead is still warm by the time a person actually calls back.