The Third Email Should Have Been a Phone Call
You sent the quote on a Tuesday. Followed up Friday. Sent a gentle nudge the following week with a subject line you rewrote four times. Nothing. The deal isn't dead — you know it isn't, because they opened the quote twice on a Sunday night — but you've run out of moves that feel appropriate. So the thread goes quiet, and in three months you'll find it while cleaning out your inbox and feel slightly sick.
The missing move was a phone call. You knew that at the time. You didn't make it because there were eleven other threads in the same condition, and calling all eleven meant losing an afternoon to voicemail.
Voice Didn't Stop Working. It Stopped Scaling.
Small businesses didn't abandon the phone because it converts badly. They abandoned it because it's the only channel where the cost per attempt never falls. An email to 40 people costs the same as an email to 4,000. A call to 40 people costs 40 calls. So the channel with the highest intent signal became the channel nobody had time for, and everything got funnelled into email, where a follow-up is cheap and therefore ignorable.
AI voice changes the arithmetic, not the etiquette. A call that asks one clear question — are you still looking at this, and would next week be better? — and records the answer is not a cold pitch. It's the thing you would have done yourself if you had four more hours in the week. The value isn't that a machine talks. It's that the attempt actually happens, on the day it's useful, for every thread that deserves one.
The risk is obvious, and worth saying out loud: voice is the most intrusive channel you own. Used as a volume play, it will burn goodwill faster than any bought list ever did. Which is why the interesting question isn't whether you can make the calls. It's who you call.
Call the People Who Already Did Something
This is where most voice tooling falls apart. It arrives as a separate product with its own contact list, so the only list you can hand it is everyone. And when the only available segment is everyone, voice becomes robocalling with better grammar.
On Growth7, voice is a channel on the same audience as email, SMS and social — which means it can be pointed at behaviour instead of a spreadsheet. The people who opened the quote twice. The people whose engagement score climbed after the last campaign and then went flat. The contacts sourced from Maps or LinkedIn who clicked once and never replied. Those are small, specific lists, and they're the only lists where a call is defensible.
That sequencing matters more than the script:
- Email and SMS do the cheap, broad work of finding out who's paying attention
- Engagement scoring tells you who moved
- Voice gets spent on the short list, where the cost per attempt is worth it
Run it the other way — voice first, to everyone — and you've built a nuisance.
The Outcome Has to Land Somewhere
A call that ends in someone's notebook is a call you paid for and threw away. Because voice sits on the same audience as everything else, the result of the conversation belongs to the same contact record: the person who said "call me in March" gets suppressed from the current campaign and surfaces again in March. The person who asked for the proposal again triggers the email. The person who said no comes off the list entirely, across every channel, not just the one they said it in.
And you can see all of this from your phone. That's not a small detail for an owner who spends the day on site — the approval queue, the campaign results and the calls that need a human follow-up are the same three things you'd check at a desk, in the app, between jobs.
The follow-up you never made wasn't a discipline problem. It was a cost problem you'd stopped noticing. Fix the cost and the discipline gets a lot easier.