How to cut repeat tenant calls with phone automation

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

Property management phone volume is rarely diverse. It's a small set of questions (office hours, rent due dates, how to submit a maintenance request, parking rules, lease renewal timing) asked over and over by different tenants. That repetition is exactly the kind of problem worth automating, but the right fix depends on how much of your call volume is genuinely routine versus how much needs a person who knows the building.

Start with a shared FAQ, it's free and it's the obvious first move

Before building anything, write down the actual questions your team answers most, and put short, correct answers somewhere every team member (and ideally tenants themselves) can find them: a pinned doc, a page on your tenant portal, a printed sheet by the phone. This alone cuts a surprising amount of repeat-call time, because a lot of it is staff hunting for the answer, not the call itself. It costs nothing and takes an afternoon.

It stops being enough once the volume is high enough that even a fast lookup adds up across dozens of calls a day, or once tenants want an answer outside office hours and a static doc can't talk back.

Where a phone-answering AI agent fits

ElevenLabs' agent platform (ElevenAgents) builds a voice agent from a written script and a set of rules, then connects it to a phone number so it can answer calls directly, any time of day. Pricing runs by plan: Pro is $99 a month with roughly 1,238 minutes of agent conversation included, Scale is $299 with about 3,738 minutes, and Business is $990 with about 12,375 minutes; usage past that allowance runs close to $0.08 a minute on paid plans, with the underlying language model billed separately. You can see ElevenLabs' current plans and what each includes before deciding which tier fits your call volume.

The part that matters for a property management use case specifically is the knowledge base: you can load your current policies, office hours, fee schedules and FAQ documents (PDF, Word, plain text, Markdown, HTML, EPUB, or straight URLs, up to 20MB per file) and have the agent answer directly from them rather than from a hand-written script alone. Smaller knowledge bases get used directly in the agent's working context; larger ones get chunked and retrieved as needed, with the total amount you can index scaling by plan, from 1MB on Free up to 1GB on Business and Enterprise. The general method is in building an AI that answers from your documents.

Handling maintenance requests specifically

Maintenance calls split into two very different kinds, and treating them the same is the most common mistake property managers make when automating this. "Is my maintenance request still open" is a status question, routine, low-risk, and genuinely well suited to an agent, as long as it can check a real ticket status rather than guessing. "My ceiling is leaking right now" is an active-problem report that needs a person's judgment on urgency, even if the underlying issue sounds mundane. Write two separate paths into the agent's rules: status checks on an existing ticket number get handled directly (once the agent is connected to your actual maintenance system, not before), and anything describing a new or ongoing problem gets a sympathetic acknowledgment and an immediate handoff, with a message logged for the right person to call back. Don't let the agent attempt to triage severity itself; a tenant describing "a little water" and a tenant describing a burst pipe use similar language, and the cost of guessing wrong is high enough that it should never be automated. For the after-hours side, see an AI receptionist for property management after-hours maintenance calls.

A realistic time-savings estimate

It helps to put a number on this before committing to a build. If your team currently spends, say, 3 minutes per call on the handful of truly routine questions (hours, rent due date, how to pay, parking rules, lease renewal timing), and those make up 60% of your inbound call volume, automating just that slice frees up a meaningful, specific chunk of staff time every week, not a vague "efficiency gain." Run this math on your own numbers before building anything: count a representative week of calls, tag each one as routine or not, and multiply the routine share by your average call-handling time. That's the actual time you're buying back, and it's a far more convincing number to bring to a budget conversation than "this should save time." Rent notices are a related case, covered in handling rent reminders and renewal outreach.

Step by step

1. List your top 10 to 15 tenant questions and write a short, current answer for each, since this becomes the agent's source material.
2. Try the shared FAQ approach first for a couple of weeks and track how much it actually reduces call volume.
3. If repeat calls are still high, load the same answers into an ElevenAgents knowledge base rather than starting from scratch.
4. Write one explicit rule: any topic that's building-specific, safety-related, or legally sensitive (an eviction question, a habitability complaint, anything involving an active dispute) gets handed to a person immediately, no attempt at an automated answer.
5. Run the agent alongside your existing phone line for a couple of weeks, check the call transcripts weekly, and update the knowledge base whenever a policy changes.

Practical tips from actually setting one of these up

• Stale documents cause more problems than missing ones. An agent answering confidently from an outdated fee schedule or an old office-hours doc is worse than it saying "let me check and call you back," because it sounds authoritative either way. Put a recurring reminder on your calendar to review the knowledge base every time a policy changes, not just at onboarding.
• Keep the regulated topics out of the automated path entirely. Anything touching habitability, eviction process, or a legal dispute should never be something the agent attempts, even cautiously. Write that exclusion into the agent's instructions explicitly rather than assuming it will naturally defer.
• Multi-property portfolios need separate knowledge bases, not one shared one. Parking rules, amenity hours and fee schedules often differ building to building. If you manage more than one property, keep each building's facts distinct in the setup so the agent doesn't accidentally give one tenant another building's rules.
• Track changes to the knowledge base the same way you'd track changes to a lease template. Per ElevenLabs' own changelog (October 5, 2026), agent deployment history is now queryable, meaning you can see what changed and when. Use that to your advantage: whenever you update a fee schedule or policy document, note the date in your own records too, so if a tenant disputes something the agent told them, you can check exactly which version of the policy was live at the time.

One more thing worth deciding up front: who owns keeping the knowledge base current. In practice this quietly becomes nobody's job unless it's explicitly assigned, since updating it isn't anyone's daily task the way answering the phone is. Assign it to whoever already owns the master policy documents, and make updating the agent's knowledge base part of the same workflow as updating the tenant portal or the printed office sheet, not a separate step someone has to remember.

What I'd tell a property manager weighing this

If your team is answering the same handful of questions dozens of times a week, the fix is almost never more staff, it's making the correct answer available without a person in the loop for the boring 80%. Where I'd push back is on the instinct to automate everything at once. Start narrow: office hours, maintenance request status, how to pay rent. Those are low-risk and genuinely repetitive. Expand only once you trust the transcripts you're reviewing, and keep anything that touches a tenant's legal standing or safety firmly on the human side, permanently, not as a phase-one limitation you plan to remove.

The other thing worth saying plainly: tenants notice when they're talking to an agent that actually knows their building versus one giving generic answers. A property management use case has a real advantage here over a lot of other after-hours phone automation, because the questions are genuinely bounded and repetitive in a way that's easy to document completely. That's exactly the setup where an AI agent performs well; it's a much harder problem for a business whose calls are unpredictable and varied, so don't assume the same approach transfers cleanly to every other part of your operation. A wider tool list is in best AI tools for property managers.

How I'd approach this

The setup above gets you most of the way on ElevenLabs' own dashboard without writing code. Where this usually needs custom work is connecting the agent to your actual property management software, so it can check a real maintenance ticket's status or a real tenant's account rather than only answering from static policy documents. That's a build on top of the agent platform, pulling live data in instead of relying on documents that go stale between updates, and it's the kind of integration I build for property teams who've outgrown the FAQ-only version. I'd rather spend an hour on a discovery call confirming the routine-versus-emergency split is drawn correctly for your specific portfolio than build the integration first and find the gaps after a tenant has a bad experience.