Try Calling Clara

Scan the QR code or tap to call and hear how Clara answers the phone.

What actually happens when an AI answers your business calls

Published 7 min read AI & Business Written by Shani Sofer
What actually happens when an AI answers your business calls

The pitch for AI receptionists is simple enough: instead of your calls going to voicemail when you can’t answer, they go to an AI that has a conversation with the caller, asks relevant questions, and sends you a summary. You get back to callers with context instead of guessing what a garbled voicemail was about.

We wanted to understand what happens beyond the pitch, so we spoke to three business owners who’ve been using Clara for several months.

Mark runs a plumbing business in Wales

He has two junior plumbers working alongside him. Most of their work comes from local referrals and Google, which means the phone rings constantly with a mix of emergencies, routine enquiries, and people who just want a rough price.

For years Mark answered every call himself. As the workload grew, the familiar pattern set in: the phone would buzz while he was on a job, he’d let it go, and by the evening he’d have a pile of voicemails to work through. “You’d listen to the message, try to understand what the problem was, then call them back and hope they still needed a plumber,” Mark told us. “Sometimes they did. Sometimes they’d already sorted it.”

His setup with Clara is straightforward. When someone calls, Clara asks about the type of problem, where the property is, and whether the job feels urgent. It doesn’t try to quote prices or make bookings. Mark wants those conversations himself. What he gets is a summary of who called and what they need, roughly sorted by urgency. “Before Clara, the phone was something you had to catch up with later. Every single day,” he said. “Now everything’s already there when I look.”

Brad runs a mobile mechanic business

His situation is different from Mark’s in that he works completely alone, always on the road. He could be servicing a car on someone’s driveway in the morning and diagnosing a warning light in an office car park by lunch. His relationship with his phone was stressful in a specific way.

“You’d finish a job, check your phone and see a handful of missed calls. Then you’re trying to remember who said what and which job is worth calling back.”

Brad’s setup captures the vehicle details, the problem, and the caller’s location. That location question turns out to matter more than you’d expect, because Brad covers a specific area and if someone’s too far away he can’t take the job. Before Clara, he’d call back, have a five-minute conversation, and only then find out they were in the wrong postcode. Now that gets filtered before he picks up the phone.

What Brad talked about most, though, wasn’t efficiency. It was relief. “It’s taken away that feeling that I need to pick up every call.” Anyone who’s been a sole trader will recognise what he means. There’s a persistent, low-grade awareness that every unanswered ring might be money walking away. You’re never fully concentrating on the job in front of you because part of your brain is monitoring the one you might be missing.

Jess runs Shine & Sparkle, a cleaning company

Her problem was different from Mark’s and Brad’s. She wasn’t missing calls. She was spending her mornings answering the same questions from new enquiries, over and over: how big is your house, what does a clean cost, can you do Wednesdays. The conversations weren’t hard but they were repetitive and they were eating into time she needed for scheduling, managing her team, and talking to existing clients.

“I wanted people to get a helpful response straight away,” Jess said. “But I also realised I was spending a lot of time repeating the same information.”

Her Clara setup goes further than what Mark or Brad do with it. Clara asks about property size, number of bedrooms, type of clean, and preferred days. It also gives callers a rough price estimate, which is something Jess chose to enable because her pricing is formulaic enough that a ballpark works. She found that when people already knew the rough cost before she followed up, the conversation was shorter and they were more likely to book. They’d already done the mental maths and decided it was worth pursuing.

“It still feels like our business. Customers get the answers they need straight away, and when I speak to them afterwards we can move the conversation forward.”

About voicemail

The reason AI receptionists are gaining traction is not that they’re exciting technology. It’s that voicemail doesn’t work.

Invoca found that the average small business misses about a quarter of inbound calls. Of those missed calls, almost nobody leaves a voicemail. The number is less than 3%. People hang up and try somewhere else.

This makes sense if you think about it from the caller’s side. You’ve got a leak, or your car is making a noise, or you want to book a clean before the in-laws visit. You call a number and reach voicemail. Now you’re being asked to explain your problem to a machine with no indication that anyone will listen to it, let alone call you back today. Meanwhile you’ve still got the leak. So you hang up and try another number, because there are always more numbers.

Clara captures these callers because it’s a conversation. Someone asks you a question, you answer it, they ask a follow-up, and by the end you’ve given your details and described what you need. The completion rate is much higher than voicemail for this reason alone.

What it won’t do

Clara doesn’t give quotes for trades where pricing depends on seeing the job. If someone asks a plumber how much it’ll cost, Clara explains that pricing depends on the specifics and takes their details. Jess’s cleaning setup is an exception, because her prices are straightforward enough that estimates help.

It doesn’t book appointments or access calendars. It captures enquiries.

It doesn’t make things up. If a caller asks whether the business does commercial work and that’s not something Clara has been told about, it says it’ll pass the question along. This is probably the most important thing about how it’s built. Getting creative with answers would be worse than not having the tool at all.

And it says what it is. Clara introduces itself as an AI at the start of every call. No attempt to pass as human. We wrote separately about how callers actually react to that, but briefly: most people are fine with it. They were going to get voicemail anyway.

The thing people kept mentioning

We expected the conversations with Mark, Brad, and Jess to be mainly about leads and revenue and conversion rates. Some of it was, but the thing all three of them brought up unprompted was more personal.

Mark said his evenings are different. He used to sit at the kitchen table working through voicemails after a full day on jobs. Jess said she didn’t realise how much of her morning the phone was consuming until it stopped consuming it. Brad used the word “feeling” twice in about ten minutes and both times he meant the weight of knowing you’re probably losing work while you’re doing work.

It’s hard to put a number on that, and we’re not going to try. But if you’ve spent time agonising over callbacks you keep meaning to make or sat in a van at the end of a long day scrolling through missed calls, you probably already know what it’s worth.

All posts

Keen to hear how Clara sounds?

Call the live demo and hear how Clara would answer enquiries for your business.

No signup required — just call and listen.