There's a version of the same question that comes up in almost every first call, usually a few minutes in, once the demo has landed and the business owner is starting to picture it live on their own phone line. It rarely arrives as a direct question. It's usually something more like, half-joking: "I just don't want people to think we've gone cheap on them." That sentence, in one phrasing or another, is the most common thing I hear in the first five minutes of a sales conversation, and it deserves a better answer than the one we used to give.
The question underneath the question
It isn't really a question about AI. Nobody who upgrades their waiting room, or redoes their website, or invests in nicer packaging gets asked whether it'll make them look cheap — those read as investment, on sight. Automating the phone line lands differently, because a lot of business owners' mental model of automation was built by decades of bad IVR menus and outsourced call centres that read, unmistakably, as cost-cutting. Underneath the joke, they're not asking about AIVA's accuracy. They're asking whether picking up this particular tool tells their customers the business stopped caring enough to have someone answer. It's one of a handful of fears we hear before anyone gets to the pricing conversation, and it's usually the first one voiced, even if it's rarely the first one meant.
Why it's not a silly question
I take this one seriously, because the underlying read isn't wrong, historically. A lot of automation genuinely has been a cost-cutting signal, deployed by businesses optimising for the lowest possible service cost rather than the best possible experience, and customers have learned to detect that and resent it. If AIVA behaved that way — slow to actually help, evasive, obviously built to deflect rather than resolve — the fear would be justified, and no amount of positioning from us would fix it. A fair amount of the product is built around exactly that worry, designing for the caller who doesn't trust it yet rather than the one who's already sold, on the theory that if it holds up for someone skeptical, it holds up for everyone else too.
The fear isn't "is this AI." It's "does using this make me look like a business that stopped caring." Those are different questions, and they have different answers.
The answer we used to give, and why we stopped
For the first several months of selling AIVA, my answer to the cheapness question was reassurance: don't worry, it sounds incredibly natural, most people won't even realise. That answer is technically true — voice quality has gotten good enough that plenty of callers don't consciously clock it — and it was still the wrong answer, because it accepted the premise underneath the question instead of addressing it. If the actual worry is "will this make me look like I stopped caring," then "no one will notice" is a strange thing to offer as comfort. It's promising to hide the decision, not defending it. And promising to hide something is itself a small tell that you think there's something worth hiding.
I noticed the answer was wrong the way you usually notice these things — a customer took it, deployed AIVA, and then asked us, a few weeks later, whether he should tell his regulars it wasn't a person answering anymore. He'd run the business for over a decade and didn't want to feel like he was misleading people he knew by name. That question stuck with me longer than the sale did. The honest answer was never going to be about concealment. It had to be about whether the thing we'd built was actually good, in a way that held up whether or not the caller knew.
What actually signals cheapness, and it usually isn't this
What we've watched, across a couple of years of customers, is that the signal people actually respond to isn't whether the voice on the phone is synthetic. It's whether their question gets answered correctly, without being made to wait or repeat themselves. A clinic that used to send every after-hours call to voicemail and now answers instantly, correctly, at 11 PM, doesn't read to its patients as a business that stopped caring. It reads as the opposite — one that finally became reachable when it mattered. The businesses whose customers do start to feel like corners were cut are almost always the ones where the automation is bad, not merely present — where it stalls, misunderstands, or sends people in circles. Customers punish incompetence, not automation. They just haven't always had a clean way to tell the two apart, because most of the automation they'd previously met was both.
It says even more when something goes sideways: whether the call quietly hands off to a person the moment it should, instead of looping a confused caller through a machine that won't admit it's stuck. We've written before about exactly what that handoff looks like — it's usually the first thing I show a skeptical owner, because it's the part that actually addresses the fear rather than talking around it.
It isn't the same fear in every industry
The shape of the worry changes by trade, too, in ways that took us longer to notice than it probably should have. A law firm partner's version of "will this look cheap" is really "will this look like we don't take the matter seriously" — the bar for gravity is much higher than for a salon booking a blowout. A restaurant owner's version is closer to "will people think we're too small to answer our own phone," which is almost the opposite fear — not that the business seems cold, but that it seems marginal. Our customer success lead has written about how the same literal question — "do you have anything available" — carries completely different weight depending on who's asking and why; the cheapness fear runs on the same logic underneath. The words people use to voice it are nearly identical across industries. The stakes underneath them aren't. Answer every business's version of the fear with the same generic reassurance and you'll be technically right and still unconvincing to at least half the room.
Where the fear is actually right
I don't think the fear is wrong often enough to wave away entirely, and I'd rather say that directly than pretend every business is a good fit for this. There are businesses where the first contact really is the relationship — a bespoke tailor with forty regular clients, a family-run funeral home, a solo therapist's intake line — where a meaningful part of what a customer is paying for is the fact that a specific person picks up. Automating that first call doesn't just risk looking cheap. It might actually be the wrong call, not a marketing problem to position around. We say this out loud when we hear it, even though it costs us the sale in the room. A high-volume clinic fielding the same handful of questions all day and a solo consultant with a dozen clients who all expect to reach her directly are not the same problem, and they shouldn't reach for the same tool just because both problems involve a ringing phone.
The honest trade a business is actually making
So the honest version of the answer, the one we give now instead of a vague reassurance, is this: you're not choosing between "cared for" and "automated." You're choosing between a phone that sometimes goes unanswered because a person can only be in one place at a time, and a phone that's always answered by something that isn't a person but is consistently competent about it. Whether that reads as cheap or as attentive, to your specific customers, depends entirely on how well the second option performs — which is a claim you get to verify yourself, on real calls, before deciding it's true. Testing it costs nothing to start, which is deliberate — this shouldn't be a decision anyone has to take on our word.
That's the answer I give now. It's longer than "don't worry, it sounds great." It's also the one that still holds up by the third week someone's actually using it — because by then, the only thing that matters is whether their customers got taken care of, not whether the voice on the line was human. If it wasn't, and the calls still went well, nobody's business looked cheap. It looked reachable, which turns out to be worth considerably more.