A prospective customer told us last month that he wanted to "add AI" to his clinic before a competitor did. Not solve a specific problem. Not fix a specific bottleneck. Add AI, the way you'd add a coat of paint — a feature you acquire, not a decision you make. I understood exactly what he meant, and it's the sentence that worries me most in almost every sales conversation we have.
The one-sentence pitch that worries me
AIVA, or any chatbot, or any voice agent, is a tool. It answers questions and books appointments extremely well, at volume, at any hour, in whatever language a customer speaks. What it is not, and what we've never claimed it to be, is a strategy for how you want to treat your customers. Installing a tool and having a strategy are different activities, and businesses that skip the second one and go straight to the first usually end up disappointed — not because the tool failed, but because nobody decided what the tool was supposed to achieve.
I say this knowing it costs us sales in the short term. "Add AI" is an easier purchase to make than "decide how you want every customer interaction to go, then configure a system to deliver that." Only one of those actually works past week three.
What a strategy actually requires
A real customer experience strategy is a set of decisions, most of which have nothing to do with AI at all. What should a customer hear in the first ten seconds of a call? Which questions deserve an instant answer and which deserve a callback? When is fast better than thorough, and when is the reverse true? What tone matches your brand — brisk and efficient, or warm and unhurried? What happens when someone is upset, confused, or asking for something you don't offer?
Concretely, the businesses that get the most out of setup tend to arrive with answers, or at least strong opinions, on a shorter list than that one: what the greeting should sound like, which two or three situations absolutely must reach a person, and what "done well" sounds like to them on a call they'd be proud of a customer hearing. That's a much smaller ask than "write a customer experience strategy document," and it gets you most of the value.
None of these are AI questions. They're business questions every business already answers, deliberately or by accident, through however its phone currently gets picked up. The businesses that do well with AIVA are the ones who'd already thought hard about these questions, or who use the process of configuring AIVA to finally think about them for the first time. The businesses that struggle are the ones who hand over a service list and a set of hours and expect the software to invent the missing decisions on their behalf.
Two businesses with the same service and the same call volume can want opposite answers to nearly every one of these questions, and both can be right for their own customers. A high-end dental practice usually wants a caller talked through the process slowly, with reassurance built into the pacing, because a chunk of that call is nervous people putting off a decision they don't want to make. A tyre and puncture shop wants the opposite — the caller already knows exactly what they need, and every extra sentence between "hello" and "yes, we have that in stock" is friction, not warmth. Neither business is more "customer-focused" than the other. They've just correctly read what their own customers are actually there for, which is precisely the judgment call a generic default can't make on a business's behalf. Writing this down as FAQs a system can actually use is a smaller, more concrete version of the same exercise, and it's often where businesses discover they'd never quite agreed internally on the answer either.
A chatbot can execute a strategy extremely well. It cannot generate one for you. That part is still your job.
"Isn't figuring this out just more work?"
The pushback I hear most, and the fair one, is that this sounds like extra homework on top of an already busy week — the whole appeal of automation was supposed to be doing less, not sitting down to write a customer experience philosophy before you're allowed to use the product. I take that seriously, because for a genuinely busy owner, "go think hard about your brand voice first" is a real cost, not a rounding error.
The honest answer is that the thinking mostly isn't extra — it's work that already needs doing, just usually done implicitly, badly, and inconsistently across whoever happens to answer the phone that day. Most small businesses have never once written down what a great call sounds like; they just have a sense of it that lives in one or two people's heads and evaporates the moment that person is out sick or moves on. Setup doesn't add that work. It just moves it earlier and makes it explicit, which is uncomfortable for about an hour and then pays for itself every week after. The businesses that skip it don't actually avoid the work — they just do it later, reactively, one awkward call at a time, after a customer has already had the mediocre version of it.
What happens when nobody decides
We can usually tell, within the first month, which customers skipped the strategy step. The calls aren't wrong, exactly — the hours are right, the prices are right, the booking goes through. They just feel slightly generic, like a competent stranger rather than someone who works there. That gap doesn't always show up as a complaint. More often it shows up as quiet disengagement: a business that deployed AIVA, never revisited the configuration, and slowly stops feeling like it's theirs. It's part of why we built out a dedicated customer success function well before we built a sales team — someone needed to be the one calling these customers back in the first few weeks, not to sell them anything, but to ask the strategy questions they'd skipped the first time.
The businesses we worry about aren't the ones who get the setup call wrong the first time — everyone does, a little. It's the ones who never come back to adjust it. A tone that was fine at launch can go stale as a business changes what it sells or who it's serving, and a strategy, unlike a plugin install, isn't a one-time decision. It's closer to a hire you keep managing. The businesses that treat it that way — revisiting the configuration every few months the way you'd check in with any employee — are consistently the ones whose calls still sound like them a year later.
The businesses that get this right
The clearest pattern from two years of customers: the ones who treat AIVA's setup like hiring and training a new senior receptionist — writing down how they want things handled, correcting it when the tone's off, deciding explicitly what should escalate — end up with something that feels like an extension of the business. The ones who treat it like installing a plugin end up with something that technically works and never quite fits.
Same software, wildly different outcomes, and the difference has nothing to do with the AI. If you want a concrete starting point rather than a philosophy, our guide to piloting AIVA properly is closer to a checklist than an essay, and it's the thing we now point people to before the sales call even ends.
What we tell people who just want the switch flipped
We still sell to the guy who wants to "add AI" before his competitor does. But the first thing we ask him now isn't his call volume or his booking system. It's what he wants a customer to feel, thirty seconds into calling his business, that they don't feel today. If he doesn't have an answer, we tell him to go figure that out first — with or without us. The software is the easy part. Deciding what you're actually trying to build was always going to be the hard part, and no amount of AI changes that. If it helps to see it working before committing to any of this, the free tier exists for exactly that reason — figure out the strategy question on a real, live line before you decide it's worth solving properly.