Chatbots vs Voice AI: Are Businesses Solving the Wrong Problem?

There is a question I get asked almost every week by business owners and senior leaders. It comes in slightly different forms, but the essence is always the same: "Should we get a chatbot?"

My answer, increasingly, is: "What are you actually trying to solve? Only then can you consider whether a chatbot is the right way forward for your business. "

That small shift in thinking matters more than most people realise and is key whenever you are considering adopting or enhancing AI in your business. It's often not as simple as Chatbots vs Voice AI.

Somewhere along the way, between the ChatGPT explosion, the wave of AI vendor pitches, and the LinkedIn articles declaring that every business needs to embrace AI right now, a curious thing happened. The technology became the destination rather than the vehicle.

Businesses stopped asking what problems they needed to solve and started asking which AI tools they should deploy; and then every other narrative was about ‘how to get the most out of Claude / ChatGPT / Gemini…’ (rinse and repeat!). Frankly, it gets a bit tedious!

Chatbots have become the default answer, and it is worth asking whether that default still serves us well.

Why Chatbots Have Become The Default AI Adoption Strategy

Let me be clear, I am not dismissing chatbots. ChatGPT fundamentally changed what people believe is possible. It democratised access to AI in a way that nothing had before. Businesses that have deployed well-designed conversational interfaces have genuinely improved customer service, reduced response times and scaled support functions without proportionally scaling headcount.

However, there is a difference between a technology being transformative and it being the right answer to every question. I strongly believe the gravitational pull towards chatbot-led AI adoption is causing many organisations to miss better opportunities and sometimes far more commercially valuable ones.

The deeper problem with the chatbot fixation is that many users (typically customers) are acutely aware they are talking to a machine. Annoyingly, the other day I conversed with a ‘chatbot’ that purported to be a real human (lovely picture of them looking all smiley and helpful), but it was a ruse – definitely a chatbot masquerading as a smiling, ‘can’t do enough to help you’ human being. It didn’t take long before it put me through to a person, and the engagement was all the better for it.

As with that example, the interaction feels transactional. It carries more than a faint sense of artificiality. I breathed a sigh of relief when I spoke to a human who could answer my questions. Much preferable to typing your query and getting a response, with a constant low-level cognitive effort involved in formulating requests, interpreting outputs, and deciding whether to trust what comes back. In many use cases, that works fine;  but for others, particularly where a customer is anxious, time-pressed, or simply wants something resolved quickly, it creates friction that the technology itself has introduced.

Which brings me to a technology that has been maturing in the background. One that most businesses are barely thinking about. And one that, for certain commercial problems, may be more valuable than any chatbot deployment.

How Voice AI Has Evolved Beyond Traditional Chatbots

Cast your mind back to around 2017. Amazon Alexa had just landed in living rooms across the country, and there was a genuine sense of wonder about it. You could speak to a device, and it would respond. It felt like the future…

…and then, within about three minutes of actual use, the frustration began.

Commands were misunderstood. You had to speak slowly and precisely, as if giving dictation to someone who did not quite follow. You repeated yourself. You rephrased. You found yourself adopting a strange, clipped, robotic cadence just to be understood. Conversations were not really conversations at all; they were structured queries dressed up as audio. Alexa was impressive in principle and deeply limited in practice (I’m not singling Alexa out for specific criticism here!), they all do that.

That experience shaped how many people still think about voice technology. It planted a scepticism that has proved surprisingly sticky, even as the underlying technology has moved on dramatically.

The Lessons from Early Voice Assistants

However, the gap between Alexa in 2017 and AI-powered voice in 2025 is not incremental. It is categorical. Large language models, real-time speech synthesis and advances in natural language understanding have combined to produce voice interactions that feel qualitatively different. Responses that adapt to context.

Why AI Voice Technology Has Changed

Conversations that tolerate interruption, topic shifts and human-sounding ambiguity. Voices that do not sound like a GPS unit reading from a script.

The technology has changed. The perception among many business leaders has not caught up.

The Business Cost of Missed Calls and Poor Customer Experience - the battleground between Voice AI and Chatbots.

Let me paint a picture familiar to many business owners.

Your phone rings. Nobody picks up. The caller (a potential new customer) does not leave a voicemail. They call your competitor, who answers, and the job is theirs…or they go to your chatbot…it can’t provide a response the potential new customer is looking for…the result is the same. They call your competitor, and you lose the business.

This happens dozens of times a week in businesses across the country, and it is almost never treated as the financial problem it is. Missed calls are written off as an operational inconvenience rather than a revenue leak, but consider the numbers for a moment.

Calculating Lost Revenue

Take a business receiving 100 calls per week. That is not unusual — a busy trades company, a professional services firm, a healthcare provider, an estate agency. If 20% of those calls go unanswered — again, not unusual, particularly during peak hours, lunch periods or when staff are with existing customers — that is 20 missed calls a week. Over a year, more than 1,000 missed enquiries.

Now apply a conservative average customer value of £500. That is potentially £500,000 in annual revenue opportunities that simply evaporated because nobody picked up the phone. At £1,000 average customer value, you are looking at seven figures. Even if only a fraction of those callers were ready to buy, the numbers are significant enough to warrant serious attention.

Most businesses do not track this. They cannot tell you how many calls went unanswered last Tuesday afternoon, let alone what those callers went on to do. The cost is real, but it is invisible — and invisible costs rarely make it onto the boardroom agenda.

The Hidden Cost Most SMEs Ignore

This is precisely the kind of operational problem that AI-powered voice technology is now positioned to address. It doesn’t do this in a gimmicky way, nor does it replace genuine human service. But as an intelligent, responsive first point of contact that ensures no enquiry falls through the gap.

Why Voice AI Is Becoming A Valuable Tool For SMEs

I recently spoke with Pierre, the founder and inventor of BAPCCO, to understand why he believes voice AI is entering a genuinely new era. BAPCCO has been working on intelligent voice solutions for business, and the conversation was one of the more grounded discussions I have had about AI in recent months; focused less on what the technology can do and more on what businesses need.

A Conversation with Pierre from BAPPCO

You have been building voice AI solutions for some time. Why does it feel different now?

"Honestly, the state of phone answering in business and the public sector right now is pretty abysmal. Long queues, 'your call is important to us,' long and unclear menu systems, and those early attempts at voice recognition that would transfer you to the Bristol ward when you asked to speak to Dr Drake. That low bar is part of why the timing feels right.

But the bigger shift is that the technology has genuinely caught up with the promise. I've been working with voice AI for around two years, and the changes in that time alone have been significant. When I do live demos, I always get the same reaction. Someone volunteers to talk to our AI receptionist, not quite knowing what to expect, and within a minute or two I see it on their face: 'WOW. That felt like an actual conversation.' Every time.

What's changed is that it no longer just understands what you're saying. It understands what you mean. It reads context, not just words. That's the real shift. And because the technology has matured so dramatically, we can now deliver that quality at a price point that surprises people. They come in expecting it to be expensive. It isn't."

Why do you think businesses consistently underestimate the cost of missed calls?

"This is a problem I've been wrestling with for years, long before BAPPCO. My background is in Quality Assurance and Governance and trying to convince a CEO to invest in either is the same challenge. You're selling a 'what if' scenario. What if our product quality were higher? How many more customers would we retain? What if we fall foul of our GDPR obligations? What does that cost us? The cost of inaction is invisible until it isn't.

Missed calls are exactly the same. You're asking a business owner to care about something they can't see and have never measured. But there's another layer to it that I think is underappreciated: accepted pain. Most owners I speak to know they're missing calls. They've just stopped feeling it. It's become background noise. And then you're asking them to change something, which introduces a new pain, a small one, but still a friction. So the calculation in their head becomes: why fix a pain I've stopped noticing?

The answer, for me, is always a real story. One of our customers, Roof Care 365, implemented our AI receptionist and, within weeks, landed a £20,000 contract from a caller who would otherwise have hit voicemail and moved on. That's not a 'what if' anymore. That's money that exists because someone picked up."

There is still a perception that AI voice interactions feel robotic and off-putting. Is that still fair?

"Honestly, even with state of the art voice AI today, there's still that uncanny valley moment where a caller thinks 'is that a robot?' But I think that's the wrong thing to focus on. The goal of the technology was never to sound human. It's to be helpful and as easy to use as possible. Those are two different things.

Alexa is a good example of why that distinction matters. It can be useful, but it was never really easy to use. You had to speak to it on its terms. What's changed with the latest generation of voice AI is that it can be both helpful and genuinely conversational, and when you get both of those things right, most people stop caring whether they're talking to AI or someone in a call centre.

I had a moment a few years back that stuck with me. I needed to call the DVLA, and the moment I heard it was an automated system, I thought, 'Here we go, there goes an hour of my life.' But the whole experience was quick, clear, and actually resolved my problem. I came away genuinely impressed. That surprise is what I see in every demo I run. People don't expect it to be that good. I'll deliberately try to trip the AI mid-demo, throw it something completely out of left field, and it handles it. Every time. That moment is usually when something shifts for people."

How do you think about the balance between AI and human involvement?

"The comparison people should be making isn't AI versus a human receptionist. It's AI versus what actually happens when no one picks up.

If you have a receptionist, great. But what happens out of hours? What happens when they're already on a call, or at lunch, or on holiday? The caller gets 'our offices are currently closed, please call back...' or they hit voicemail. Our AI receptionist will never be as good as a good human receptionist. But it's a hell of a lot better than voicemail.

If you don't have a receptionist, the situation is just as clear. I have an accountancy firm as a customer, small team, no receptionist. When the phone rings, one of the accountants picks up. That means dropping what they're doing to answer the same question for the twentieth time, or to tell a sales caller they're not interested. The AI handles all of that, and gives the accountants their time back. Another customer told me that since implementing our solution, every call they now take is one they actually want. Everything else is handled, messaged, or redirected. That changes the value of every single conversation they have.

I've also seen the other side of this first hand. In a previous role I was involved in a project that used AI to replace staff rather than augment them. It did not go well. I said it to the CEO at the time and I'll say it now: companies that use AI to replace people rather than work alongside them are storing up serious problems. The only way to get the best out of this technology is humans and AI working hand in hand."

How do customers actually respond when they realise they have been speaking with an AI?

"The reactions are more nuanced than most people expect, and I'll be straight about it because I've actually listened to these calls.

Callers broadly fall into three groups. A small minority, around 5% or less, hang up the moment they realise it's AI. They're not interested, and that's their right. The majority just want their question answered. If the AI does that, they're fine with it, sometimes genuinely surprised by how smooth it was. And then there's a third group who don't even clock that they've been speaking to AI, despite the agent introducing itself as 'the AI receptionist for...' at the start of the call.

The occasional caller rings back just to ask if it was AI, or to have a bit of fun with it by throwing random questions at it. Given the current noise around AI taking people's jobs, I wouldn't expect anything different. That anxiety is real and understandable.

What I'm clear about, and what I encourage all our customers to be clear about, is that it is a legal requirement to confirm you're talking to AI if a caller asks. I don't just comply with that, I agree with it. The goal was never to fool anyone into thinking they're speaking to a human. The goal is to help them solve a problem or get an answer. If we do that well, most people are fine with how it happened."

Where do you see voice AI in five years?

"I think about this generationally, because the answer looks different depending on who you're talking about.

Baby Boomers are already using voice in the best possible way: they call people. There's something almost poetic about that.

Gen X will need convincing. Give them a keyboard and they'll tell you exactly what they think on Facebook. Or X. Whatever we're calling it this week. They've earned their scepticism and the technology will have to earn their trust.

Millennials grew up watching smartphones go from novelty to necessity. They use their technology well and efficiently already. Voice AI is going to have to be genuinely better, not just different, to pull them away from those lightning fast thumbs.

Gen Z are digital natives who live on their phones. For them it'll come down to one question: is this faster? If voice AI can beat their swipe speed, they're in.

And then there's Gen Alpha, the first generation raised alongside Alexa, Siri, and whatever came after. For them, talking to technology isn't a novelty. It's just how things work.

My honest view is that in five years voice AI won't replace existing technology, it'll work alongside it. Voice for some things, screens for others, each doing what it does best. BAPPCO's goal is to be part of that journey, bringing these tools to UK businesses first, then Europe, then the world… One can dream."

AI Receptionists: A Practical Voice AI Solution For SMEs

What struck me about BAPCCO's approach is that it starts with a business problem most owners recognise immediately. The TL;DR version is: unanswered calls, customer enquiries that slip away, and inconsistent first impressions.

The Receptionist solution combines AI with professional voice actors to create interactions that feel genuinely human in quality, not deceptive. The goal is not to convince callers they are speaking with a person. It is to ensure the interaction is helpful and responsive and leaves a positive impression of the business. That distinction matters, and BAPCCO takes it seriously.

For an SME owner who cannot justify a full-time receptionist but knows their team cannot always get to the phone, this is not an AI experiment. It is a practical answer to a real operational problem.

It’s worth visiting the BAPCCO website. They have some live demos of their AI Receptionist solution that are well worth a listen. I can heartily recommend the ‘patience’ demo to fully understand just how good the solution is.

Why AI Strategy Should Focus On Business Problems First

I want to return to where we began, because it is the point I most want to leave you with.

The worst AI projects I observe do not fail because the technology did not work. They fail because the business started with a tool in mind and worked backwards to justify it. "We should have a chatbot" is not a strategy. Neither is "we should automate something."

The right questions are harder but far more productive. What enquiries are we failing to respond to? Where are we losing customers before they ever become customers? What does a poor first interaction cost us, and how often does it happen?

Sometimes a chatbot is the right answer, especially for high-volume digital enquiries, for website visitors who need information at any hour, and for support functions where speed matters more than warmth. Sometimes, workflow automation is more appropriate, removing repetitive manual processes and connecting systems that do not speak to each other. However, sometimes, as I have been arguing here, a voice solution is where the real value lies, particularly for businesses where the telephone remains the primary channel for new enquiries, where first impressions carry commercial weight, and where unanswered calls represent a measurable and recoverable revenue problem.

The technology is not the answer. The answer is the answer. The technology is how you get there.

What Voice AI Means For The Future Of Customer Service

I do not usually advocate a tools-first approach, and I am not doing so here. However, for organisations that know (genuinely know) that they miss calls, struggle to respond consistently, or want to improve how they handle first contact with potential customers, AI-powered voice represents one of the clearest, most commercially grounded business cases available today.

There is also something more interesting happening here, which is worth considering for a moment. In a world saturated with text-based AI interactions, chatbots, email automation, and AI-generated content, the human voice may be the way we bring genuine connection back into technology. This isn’t because AI voice is human; it’s because voice, by its nature, is closer to how we communicate with one another than any screen-based interface has ever been, and I, for one, hope it remains that way.

The chatbot era is not over. But it may be maturing into something richer. Certainly, for the businesses paying attention, that shift represents an opportunity worth taking seriously before their competitors do.

Ready to Think More Clearly About AI?

If you are just beginning your AI journey and want to understand where the real opportunities lie for your business, start with the Talisman AI Readiness Assessment. It will give you a grounded, honest picture of where you stand and where to focus first.

If you are already embedding AI across your organisation and want to understand whether your adoption is genuinely effective, sustainable and delivering lasting value, the Fit for AI diagnostic is designed precisely for that conversation.

For more AI-related insights; read the Talisman AI FOTO ('From Obstacles to Outcomes' Series)

Either way, the place to start is the same: with the problem, not the tool.

Talisman AI Consulting helps SME leaders and operational decision-makers cut through the noise and find the AI opportunities that matter for their business. Get in touch at roger@talisman-ai.co.uk