Most leads do not die because the offer was wrong, the ad was weak, or the price was off. They die in the silent gap between the moment someone hits "submit" and the moment a human being on your team actually talks to them. For a typical NZ or Australian services business — solar, insurance, mortgage, real estate, a law firm, an accountant — that gap is somewhere between two hours and two days. By then the lead has already filled out three competitor forms and taken a call from the first one to ring.
For years the conversation around fixing this has been a single-channel argument. "We need a chatbot." "Get an SMS tool." "Hire an answering service." More recently: "Just put a voice AI on it." Every vendor pitches their channel as the answer. After running this for a few years on our own clients, I can tell you the answer is not picking one. The answer is running all three — voice, SMS and chatbot — as a single coordinated stack with one brain behind them. That is what our AI automation system is built around, and it is why we can price the way we do.
Here is the case for each channel, where each one breaks, and how they compound when you stop treating them as alternatives.
Voice AI: the only channel that wins the speed race
Voice AI is the front door. The moment a lead submits a form, the AI rings their phone — typically inside 60 seconds. There is no other channel on earth that can do this consistently at scale. A human team can sometimes pull it off for the first lead of the morning. They cannot pull it off for the lead that submits while they are on another call, in a meeting, at lunch, asleep, or simply on a different job site.
When a modern voice agent gets through, it is genuinely good. The conversation flows. It asks the qualifying questions your best salesperson would ask in the first two minutes. It explains the offer. It books the calendar. We are not in 2019 anymore — these are not robotic IVRs. Half the prospects we listen back to never realise they were talking to AI, and the ones who do figure it out tend to be impressed rather than annoyed.
So what is the catch? People do not pick up the phone.
Across the campaigns we run, 60–70% of leads do not answer the first voice attempt. That is not a voice-AI problem — that is a 2026 phone-behaviour problem. New mobile numbers, unknown caller ID, screen-first culture, "I'll call them back later" syndrome. If voice is your only follow-up channel and you cannot reach somebody on the first ring, you have lost them. Whoever can re-engage them in a way they will actually respond to wins.
SMS: where voice fails, text picks up
This is the bit that most agencies and DIY operators miss. The instant a voice call goes unanswered, the same AI should be texting the same person from the same number, in the same tone of voice, continuing the same conversation it would have had on the phone.
This is not a blast template that says "Hi, this is XYZ Solar, we tried to call you, please call us back." That is the kind of SMS that gets ignored or reported as spam. What works is a real, conversational, time-zone aware message that opens with something specific to what the lead just filled out:
"Hey Mark, this is Solar Hub — just tried to give you a quick ring about the panel quote you requested. Easier by text? Happy to send through a couple of times that suit for the site assessment."
That gets a reply. From there the AI handles the conversation as a normal text thread — answering questions, surfacing the right qualifying info, sending the calendar link, confirming the booking. It does not get tired and it does not forget to follow up two days later. For leads in different states or time zones (an Auckland brand selling into Australia, or a Sydney firm getting NZ enquiries) the SMS engine is time-zone aware so the lead never gets a text at 11pm and instantly hates you.
SMS is where the majority of bookings actually happen for the leads voice could not catch. It is the channel that turns "I missed the call" into "I'm in the calendar for Tuesday."
AI chatbot: the inbound side of the same brain
Voice and SMS handle outbound — the moment after the lead has raised their hand. The chatbot handles inbound. Somebody lands on the website at 10pm, has a question, and is not ready to fill out a form yet. Or they message the Facebook page on a Sunday. Or they hit the business on WhatsApp because that's what they use for everything else in their life.
If those visitors get silence, or worse, "We'll get back to you in 1–2 business days," they're gone. A properly built AI chatbot — same knowledge base as the voice agent, same qualifying questions, same calendar — meets them where they are. Web widget, Facebook Messenger, WhatsApp. One brain across all three surfaces. It answers the actual question they asked (not a canned FAQ), qualifies them inline, and books the appointment directly. No human in the loop until the appointment shows up in the calendar with the qualification notes attached.
The chatbot's job is not to replace the website's "Contact us" form. It is to catch the inbound traffic that would otherwise bounce because nobody was available, and to handle the channels (Messenger, WhatsApp) where forms aren't a natural fit at all.
The compounding effect: one ad spend, three engagement attempts
This is the part that changes the economics. Consider the same lead viewed two different ways.
Single channel (just voice): Lead submits → voice calls → 65% of the time they don't pick up → lead is essentially lost unless a human chases them later, which they usually don't. Ad spend has effectively been wasted on the unreachable two-thirds.
Three-channel stack: Lead submits → voice calls within 60 seconds → if no answer, SMS opens the same conversation by text within minutes → meanwhile, if the lead comes back to the website to look around, the chatbot greets them and offers to book directly. You now have three different shots at engagement off the same ad spend, in the windows the lead is most receptive in.
This is not theory. Solar Hub NZ ran the single-channel approach for years. Their lead-to-appointment ratio sat around 44%. After we layered voice + SMS + chatbot as a single stack on top of their existing ad campaigns, lead-to-appointment moved to 88% in 9 months. The qualified pipeline grew 3.8×. Cost per install dropped 41%. The ad spend didn't materially change — the engagement stack on top of it did.
The lift comes from the simple fact that humans engage on different channels at different moments. Some pick up the phone. Some only reply to texts. Some come back to the website three hours later and want to chat before committing to a booking. If you only run one of those channels, you only catch a slice of the demand you already paid for.
Why this finally makes "pay per appointment" possible
For a long time the honest pricing models for lead handling were either retainer ("pay us $3k–8k a month and we'll do our best") or per-lead ("pay $80–200 per form fill, hope they qualify"). Both put the risk on the business owner. Retainers get paid whether or not anything books. Per-lead pricing rewards volume of bad leads as much as good ones.
When voice, SMS and chatbot are running as one stack, with the same qualifying questions across all three and a shared custom CRM workspace tracking every interaction, something changes: the booked appointment becomes a reliable enough output that we can price against it. That's why our pricing is a single line — $150 NZD + GST per confirmed appointment on your calendar. No retainer. No setup fee. No per-lead bill for tyre-kickers. If the appointment doesn't land in your calendar, we don't get paid.
That pricing only works because the multi-channel stack lifts the conversion rate enough to absorb the cost of the unreachable leads. With a single-channel system, the unit economics just don't run. With three channels working off one brain, they do. That is the entire reason we built the system the way we built it — and the reason our qualification methodology is so tightly tied to the channel stack.
What this looks like in practice for your business
If you're a solar installer, mortgage adviser, insurance broker, real estate agent or professional services firm and you're considering AI for lead handling, the question to ask the vendor is not "do you do voice AI?" The question is: "Are all three channels — voice, SMS, chatbot — running off the same knowledge base, the same qualifying questions, and the same calendar, with one CRM showing me every conversation in one place?"
If the answer is no — if voice is one vendor, SMS is a Zap, and the chatbot is a different widget that doesn't know what the voice agent said — you're going to end up with the same gaps the single-channel approach has, just with more monthly subscriptions.
The whole point of the stack is that it's a stack. Not three tools duct-taped together.
If you want to see what this actually looks like end-to-end — the campaign document, the qualifying questions, the voice/SMS scripts, the chatbot, the CRM and the dashboard — the easiest way is to look at what we shipped for Solar Hub and book a 30-minute call. We'll walk you through the system we'd build for your business and show you the unit economics on a whiteboard before you commit to anything.