The 6 Questions Your Front Desk Should Ask Every Caller, and How to Automate Them

The call that should have been a job

Picture this. A homeowner calls your business at 2:30 on a Tuesday. Your tech is on a roof, your admin is at lunch, and the call goes to voicemail. The homeowner hangs up and calls the next number on Google. That job, probably worth $400 to $1,200, is gone. You never even knew it came in.

That is the obvious version of the problem. The less obvious version happens when someone does answer. They pick up, they're distracted, they say "Sure, we can come out, what's your address?" and that's basically it. No job type confirmed. No urgency established. No budget range explored. No one asked whether the customer is a homeowner or a property manager. The appointment gets booked, the tech shows up, and half the information needed to do the job right is missing.

Both versions cost you money. The first costs you the lead entirely. The second costs you efficiency, trust, and sometimes the job itself when expectations don't match reality on-site.

The fix is not complicated. It is six questions, asked in order, every single time. The problem is that humans are inconsistent, especially when they're juggling three other things. So let's talk about what the questions are, why each one matters, and how to get them asked reliably without hiring a full-time receptionist.

Why most front desks skip the intake

It is not laziness. Most of the people answering phones at small service businesses are wearing four hats. They're scheduling, handling supplier calls, dealing with a tech who needs a part number, and trying to eat lunch. When a new call comes in, the goal becomes "get this person off the phone with an appointment booked" rather than "gather everything we need to actually serve this person well."

The result is a booking that is half-baked. You have a name and an address. You don't have the information that determines how long the job takes, whether it requires a specialist or specific equipment, or whether the customer's expectations are anywhere close to what you're going to charge.

A consistent intake script solves this. But a script only works if someone follows it. That's where automation earns its keep.

The 6 questions, and why each one matters

1. What's the problem or service you need today?

This sounds obvious, but you would be surprised how many calls get booked as "general service" because nobody asked specifically. Is it a leak, an installation, a tune-up, an estimate? The answer determines job duration, which tech to send, what parts to pull, and what to charge. Without it, you're scheduling blind.

2. How urgent is this for you?

Some customers have been sitting on a slow drain for six months. Others have water coming through the ceiling right now. If you don't ask, you have no way to triage. You might schedule your emergency call for Thursday while the customer needed someone today and already called someone else by the time you get there. Asking urgency also opens the door to premium same-day or emergency pricing, which is completely fair and customers will accept it when it's presented honestly.

3. Has this issue been looked at before, and if so, what was found?

This one saves techs from walking into a situation blind. If someone already had two companies out and neither one fixed it, your tech needs to know that before arrival. It sets the right expectations, prevents the tech from being blamed for prior failures, and helps you send the right person with the right experience.

4. Are you the homeowner, a tenant, or a property manager?

This question has billing and authorization implications. A tenant may not be able to approve work above a certain dollar amount. A property manager needs a separate invoice and may have specific vendor requirements. A homeowner is straightforward but may need financing options mentioned. One question, very different downstream processes.

5. What is the address, and is there anything about access we should know?

Address is table stakes. The access part is where most intakes fall apart. Gate codes, aggressive dogs, parking restrictions, units in apartment buildings, hours the property is accessible, whether someone will be home. A tech who can't get in wastes a trip. A wasted trip costs you labor, fuel, and the customer's trust.

6. What is the best number to reach you, and do you prefer a call or a text for your reminder?

Confirmation contact preferences matter more than people think. If you send a call reminder to someone who never picks up unknown numbers, they miss it and no-show. If you text someone who is 70 years old and barely uses their phone, same result. Asking takes five seconds. It cuts no-shows meaningfully. Industry data puts no-show rates for service appointments between 10 and 20 percent at businesses that don't confirm properly. Dropping that by half is real money.

What it actually costs you to skip this

Let's put some numbers to it. The table below compares a business running loose intake versus one running consistent intake across the same call volume.

That is a rough model, but the direction is always the same. Bad intake is expensive in ways that don't show up as a line item. It shows up as techs complaining about bad jobs, customers who feel surprised by costs, and reviews that mention "poor communication."

The real-world version of this problem at 7 PM on a Friday

Your office closes at five. A homeowner has a water heater that started making noise and they're worried it's going to fail over the weekend. They call. Voicemail. They call a competitor. Competitor answers, asks good questions, books the job for Saturday morning with an emergency surcharge. You find out Monday when you check your missed calls.

This is not a hypothetical. It is happening to most service businesses every single week.

The solution is having something that answers when you can't, knows what to ask, and captures the answers in a usable format. That is exactly what a purpose-built AI receptionist does. If you want to see what that experience feels like from the customer's side, go to our AI receptionist page, paste your own website URL into the tool, and within about 20 seconds you're talking to an AI that has already read your business and is ready to handle an intake call. No signup, no credit card, nothing installed. You just talk to it like a customer would, and you'll immediately understand why it captures more than a voicemail ever will.

How to actually automate this intake

There are a few ways to approach it, and the right one depends on your volume and how your business currently handles calls.

Option 1: Build a phone script and train your staff

Cheapest option. Print the six questions on a card and put it next to every phone. Run a 15-minute training. Follow up by listening to calls periodically. This works if you have a dedicated receptionist who is focused on calls. It fails when that person is multitasking, overwhelmed, or out sick.

Option 2: Use a call answering service

Services like Ruby or Posh will answer calls and follow a script you provide. Cost runs roughly $250 to $600 per month depending on call volume and hours of coverage. They're human, which some customers prefer, but the quality of intake varies depending on which agent picks up. They also can't always access your actual scheduling system in real time.

Option 3: Use an AI receptionist paired with your CRM

This is the setup that handles the most scenarios well. An AI receptionist answers 24/7, asks the six questions (or whatever intake you configure), captures the answers in structured form, and either books the appointment directly or flags it for follow-up. When it connects to your CRM, the booking shows up immediately with full intake notes attached. No one has to transcribe a voicemail at 8 AM and hope they got everything.

The cost comparison matters here. A part-time admin focused only on phone intake runs you $18 to $25 per hour, or roughly $3,000 to $4,000 per month for full coverage including evenings. A call answering service runs $250 to $600 per month but usually doesn't cover late nights or weekends reliably. An AI receptionist solution typically runs $100 to $300 per month and covers every hour without variation in quality.

There are cases where a human answering service genuinely wins, specifically if a significant portion of your callers are elderly or in markets where people react poorly to any hint of automation. Be honest with yourself about your customer base before committing either way.

Setting it up: a step-by-step

1. Write your six questions down in plain language, exactly as you'd want them asked out loud. Avoid jargon. "What's going on with your system today?" beats "Please describe the nature of the service request."
2. Add any business-specific follow-ups. If you do both residential and commercial, add that distinction. If certain job types require photos before dispatch, note that so the AI can ask the customer to text photos to a specific number.
3. Connect the intake to your scheduling system. The goal is zero manual transcription between "call ends" and "appointment appears on the schedule."
4. Set up confirmation outreach. A text confirmation sent within five minutes of booking, followed by a reminder 24 hours out, will do more to reduce no-shows than anything else in this list.
5. Review intake records weekly for the first month. Look for questions that customers are getting stuck on or answering in ways that aren't useful. Tweak the wording.
6. Check your no-show rate and wasted trip count at 30 and 60 days. If it's not moving, something in the intake or confirmation flow is broken.

If you want to see what a full-stack version of this looks like, including intake, CRM, website, and follow-up, the Corex modules page breaks down how the pieces fit together for service businesses. You don't have to buy everything at once. Most businesses start with the receptionist function and add from there.

FAQ

Will customers push back on answering questions before they get a price?

Some will. The framing matters. Position the questions as "I want to make sure we send the right person with the right parts so we don't waste your time." Most customers respond well to that. The ones who refuse to answer anything and just demand a price are often the hardest jobs anyway. Getting more info before dispatch filters out some of the most frustrating interactions, which is a side benefit worth mentioning.

What if a caller has an emergency and just needs someone now?

A good intake flow has a short path for emergencies. Two questions max: what's happening, and what's the address. Then a human gets notified immediately or the caller is routed to an on-call line. The full six-question intake is for standard bookings. Emergencies need a different branch, and any decent AI receptionist can be configured to recognize urgency signals and skip to dispatch mode.

How do I handle calls where the customer isn't sure what they need?

That's actually a valuable intake outcome too. "I'm not sure, something smells weird near the furnace" tells you a lot. You need a tech who can diagnose, not just perform a pre-scoped task. Log it as a diagnostic visit, set the right expectation on pricing for the service call, and make sure the tech knows it's open-ended going in. The intake question doesn't need a perfect answer to be useful.

Can an AI receptionist handle thick accents or people who are hard to understand?

Current AI voice tools have gotten significantly better at this over the past two years, but it is still a real limitation for some callers. The honest answer is that most mainstream AI receptionist tools handle standard American accents very well, and struggle more at the edges. If your customer base includes a large proportion of non-native English speakers, test specifically for that before committing. Some businesses run AI for standard calls and keep a human fallback for anything the AI flags as unclear.