Full transcript

Section 1
Yes, people are getting calls from ChatDBT, but the volume is really low right now, and the smarter move is to measure it before you bet any budget on it.

Section 2
And that’s what this video is about. I’m going to show you the real numbers, how to track AI referred calls, and what a typical home service business should do about it in the next 30 days.

Section 3
The honest answer on volume.

Section 4
you. Here’s a truth that nobody wants to say out loud. ChatGPT is sending calls, but not many.

Section 5
In our own AI tracker at Pushleads, we ran 65 checks over 30 days, and ChatGPT named us in

Section 6
five of them, and that’s five out of 65. That’s the lowest of any engine that we track, and that’s

Section 7
us, a marketing company that specifically works on AI visibility. If that’s low for us, assume it’s

Section 8
low for you too. The total call volume coming from generative AI tools is still very small

Section 9
across the country. Invoca’s Home Service Lead Conversion Benchmarks report, published in July

Section 10
of 2026, confirmed that no other AI tool besides Chat2BT is producing measurable call volume

Section 11
at all. So we’re very early, very early. But the quality is unusually high. And here’s where it

Section 12
gets interesting and why you shouldn’t just ignore this channel that same invica report looked at

Section 13
phone lead rates across every major traffic source and chat gpt calls came in at 45 phone lead rate

Section 14
paid search which most of you are already running came in at 40 so that’s a five point gap in favor

Section 15
of ChatGPT. And those ChatGPT calls converted at 41%. So when someone calls you after finding you

Section 16
in ChatGPT, they’re more likely to become a customer than a caller from most other sources.

Section 17
The volume is low, but the quality is real. Why the quality gap exists. Think about the

Section 18
difference in intent. When someone types a question into ChatGPT, they’re usually describing

Section 19
a specific problem, not something like roof or near me, but something like I have a leak above

Section 20
my bathroom after last night’s storm. Who should I call in my area? And that’s a buyer and they’ve

Section 21
already processed the problem and they’re asking for a name, not options. So when ChatGPT gives

Section 22
them your name, they already call with the decision half made. And that’s why the conversion rate is

Section 23
higher paid search callers are still shopping and chat gpt callers are often more ready to book

Section 24
the skepticism is real too before i tell you to go all in i won’t because i think you should know

Section 25
that you’re not alone if you’re skeptical about this jobbers research found that nearly one in

Section 26
five home service business owners don’t believe ai will help their business at all and that’s a

Section 27
reasonable position right now. The volume just doesn’t justify blowing up your budget. If paid

Section 28
search is working for you, keep running paid search. If your LSA leads are strong, protect that.

Section 29
AI referred calls are just a channel to watch and measure, not a channel to pivot your whole

Section 30
business around. Not yet. What measuring actually looks like. So what does it mean to measure this

Section 31
properly? There are three things you need in place. First, call tracking by source. Every phone

Section 32
number you use needs to be a track number so that you can see where each call originated if someone

Section 33
clicks a link from chat gpt to your website then calls you that shows up as a website visit call

Section 34
and you’ll see the referral source as chat gpt if your tracking is set up right second ask every

Section 35
caller how they found you so train your fund desk or your csr to ask the question how did you hear

Section 36
about us? And asking that question just takes three seconds. And it catches calls where the

Section 37
AI visit did not get attributed. Third, track it next to your other channels so you can compare

Section 38
volume, cost, and outcome in one place. A real example, Bear Restoration. Let me show you how

Section 39
this worked for the real example. Bear Restoration, based in Northwest Arkansas, gave us permission to

Section 40
share their numbers over the last 30 days october 18th to november 16th of 2026 they logged a total

Section 41
of 82 leads and that was 36 calls and 20 google lsa leads and 26 web forms every phone call was

Section 42
tracked with the source and outcome so at any point in the last 30 days they know exactly which

Section 43
calls came from which channel and which ones booked and that’s the baseline you need before

Section 44
you can evaluate a channel like ChatGPT. And if you don’t have that baseline, you don’t know what’s

Section 45
working and you definitely won’t know when ChatGP starts sending you leads. How the bear dashboard

Section 46
works. In the bear dashboard, calls sit right next to forms and LSA leads. So you’re not toggling

Section 47
between three different platforms. You’re looking at one screen that shows where the lead came from,

Section 48
when it came in, and what happened to it. And when a call comes in as a tagged AI referral,

Section 49
it doesn’t get lost in the spreadsheet it shows up in context next to

Section 50
everything else that matters because AI referred call volume is low right now so

Section 51
if you’re only checking once a month you’ll miss two or three calls that

Section 52
actually came through you’ll need to see them in real time next to your other

Section 53
channels so you can judge whether the call volume is trending up the tracking

Section 54
setup you need here’s the minimum setup I’d recommend one track phone number for

Section 55
your website one for your Google business profile and one for your paid

Section 56
search if you’re running it, and a system that asks the source question at every inbound call.

Section 57
This doesn’t have to be complicated, and a lot of call tracking platforms will let you

Section 58
record the intake call and log the source automatically. And if you’re a Pushleads

Section 59
client, this is already built into our own dashboard. If you’re not, the principle is the

Section 60
same. Every lead source gets its own phone number, and every call gets logged, and you review it

Section 61
weekly. What ChatGPT actually sees when you look it up. Here’s something that most business owners

Section 62
don’t know. ChatGPT doesn’t just search the web the same way that Google does. It’s pulling from

Section 63
the mix of its training data, real-time web browsing in some cases, and third-party data

Section 64
sources like Yelp, Angie, and other directories. So if your business information is inconsistent

Section 65
across all these platforms, ChatGPT may not recommend you or may recommend you with the

Section 66
the wrong details. Your name, address, and phone number all need to match up everywhere. And this

Section 67
is called the NAP. And this is also important for the Google Business Profile as well. So your

Section 68
reviews also matter on these platforms. The way people describe your services on review sites is

Section 69
part of what ChatGP reads when it decides who to recommend. What gets you named in ChatGPT?

Section 70
In my experience working on AI visibility, the businesses that get named most reliably have a

Section 71
few things in common. Consistent information everywhere. Recent specific reviews that

Section 72
mentioned the type of work. Not just something like great company, but they fixed our roof after

Section 73
the hailstorm and had us dry by Friday. A website that answers the questions clearly, not just a

Section 74
homepage with a phone number. And mentions on third-party sites, local news, neighborhood groups,

Section 75
industry directories. ChatGPT needs reasons to trust your name. The more specific the evidence,

Section 76
the better. Common mistakes that business owners make. The biggest mistake I see is ignoring AI

Section 77
entirely. Not because the volume is high now, but because if you ignore it now, you’ll have no

Section 78
baseline when the call volume grows. You won’t know if your three AI refer calls this month

Section 79
turned into 30 next year. The second mistake is the opposite, spending heavily on AI-specific

Section 80
campaigns before the volume justifies it. Right now, the return on that spend is just unproven.

Section 81
so measure first and spend later. The third mistake is assuming that your existing tracking

Section 82
already captures AI-referred calls. Most standard setups don’t spit this out. You have to check

Section 83
specifically whether chat GPT referrals are being recorded as a distinct source in your system.

Section 84
Have the math works for a typical setup. Picture a restoration company taking in about 50 inbound

Section 85
down calls a month. If the industry pattern from Imbica holds and 45% of AI referred calls become

Section 86
leads and 41% of those convert, then even two or three of those AI referred calls a month could

Section 87
mean one additional job. For a water damage call, that’s real revenue. The question isn’t whether

Section 88
two calls is impressive. The question is whether you’re capturing those calls, logging them, and

Section 89
and following up fast, because callers who found you through ChatGPT are ready to hire,

Section 90
and if you miss them or don’t call back, that quality advantage disappears.

Section 91
What the next 30 days should look like. Here’s a simple 30-day action plan. Week 1,

Section 92
audit your tracking setup. Can you see ChatGPT as a referral source in your analytics or call

Section 93
tracking? If not, fix that first. Week 2, audit your directory listings. Check that your name,

Section 94
address and phone number are identical on Google Yelp Angie and any other

Section 95
platform you’re listed on week 3 collect 2 or 3 reviews that describe the

Section 96
specific type of work you do ask a recent customer to mention the job type

Section 97
the location and the outcome week 4 check your website does it answer the

Section 98
questions that someone in a crisis would ask what do I do if my roof is leaking

Section 99
right now is a better page than our roofing services why this matters more

Section 100
for home services. Home services is the category where AI recommendations are more likely to have

Section 101
the most impact. When someone’s basement is flooded at 10 at night, they’re not browsing

Section 102
seven websites. They’re asking chat, GBT, or a similar AI who to call right now. And the businesses

Section 103
that show up in that answer get the call. The ones that don’t, don’t. And that’s not a theoretical

Section 104
future scenario. That’s happening now with low call volume. The volume will grow. The business

Section 105
building their AI presence now will have an advantage when it does. The businesses waiting

Section 106
for it to become obvious will be playing catch-up. Get your free AI visibility check. Here’s what I

Section 107
suggest you do today. Go to pushleads.com slash check and get your free AI visibility check.

Section 108
We’ll look at whether ChatGBT and other AI tools are finding your business, what they’re saying

Section 109
about you, and where the gaps are. You can also call us directly at 828-348-7686. And there’s no

Section 110
No cost for the AI visibility check, and you’ll walk away knowing exactly where you stand with AI search right now.

Section 111
Not where you hope to stand, where you actually stand.

Section 112
That’s a starting point.

Section 113
Measure it, then improve it.

Section 114
I’m Jeremy Ashburn with Push Leads, and I help small businesses get more leads without struggle and frustration.

Section 115
Talk to you soon.

▶ Watch it hereMore videos →

Is anyone actually getting phone calls from ChatGPT?

ChatGPT calls convert at 41% — higher than paid search — but volume is still small, making now the perfect time to build your tracking baseline. Jeremy Ashburn of Push Leads breaks down real numbers, a live client example, and a clear 30-day action plan so home service businesses can measure AI-referred leads, fix dire

Yes, people are getting calls from ChatDBT, but the volume is really low right now, and the smarter move is to measure it before you bet any budget on it. And that’s what this video is about. I’m going to show you the real numbers, how to track AI referred calls, and what a typical home service business should do about it in the next 30 days. The honest answer on volume. you. Here’s a truth that nobody wants to say out loud. ChatGPT is sending calls, but not many. In our own AI tracker at Pushleads, we ran 65 checks over 30 days, and ChatGPT named us in five of them, and that’s five out of 65.

That’s the lowest of any engine that we track, and that’s us, a marketing company that specifically works on AI visibility. If that’s low for us, assume it’s low for you too. The total call volume coming from generative AI tools is still very small across the country. Invoca’s Home Service Lead Conversion Benchmarks report, published in July of 2026, confirmed that no other AI tool besides Chat2BT is producing measurable call volume at all. So we’re very early, very early. But the quality is unusually high.

And here’s where it gets interesting and why you shouldn’t just ignore this channel that same invica report looked at phone lead rates across every major traffic source and chat gpt calls came in at 45 phone lead rate paid search which most of you are already running came in at 40 so that’s a five point gap in favor of ChatGPT. And those ChatGPT calls converted at 41%. So when someone calls you after finding you in ChatGPT, they’re more likely to become a customer than a caller from most other sources. The volume is low, but the quality is real. Why the quality gap exists. Think about the difference in intent.

When someone types a question into ChatGPT, they’re usually describing a specific problem, not something like roof or near me, but something like I have a leak above my bathroom after last night’s storm. Who should I call in my area? And that’s a buyer and they’ve already processed the problem and they’re asking for a name, not options. So when ChatGPT gives them your name, they already call with the decision half made.

And that’s why the conversion rate is higher paid search callers are still shopping and chat gpt callers are often more ready to book the skepticism is real too before i tell you to go all in i won’t because i think you should know that you’re not alone if you’re skeptical about this jobbers research found that nearly one in five home service business owners don’t believe ai will help their business at all and that’s a reasonable position right now. The volume just doesn’t justify blowing up your budget. If paid search is working for you, keep running paid search. If your LSA leads are strong, protect that.

AI referred calls are just a channel to watch and measure, not a channel to pivot your whole business around. Not yet. What measuring actually looks like. So what does it mean to measure this properly? There are three things you need in place. First, call tracking by source. Every phone number you use needs to be a track number so that you can see where each call originated if someone clicks a link from chat gpt to your website then calls you that shows up as a website visit call and you’ll see the referral source as chat gpt if your tracking is set up right second ask every caller how they found you so train your fund desk or your csr to ask the question how did you hear about us?

And asking that question just takes three seconds. And it catches calls where the AI visit did not get attributed. Third, track it next to your other channels so you can compare volume, cost, and outcome in one place. A real example, Bear Restoration. Let me show you how this worked for the real example.

Bear Restoration, based in Northwest Arkansas, gave us permission to share their numbers over the last 30 days october 18th to november 16th of 2026 they logged a total of 82 leads and that was 36 calls and 20 google lsa leads and 26 web forms every phone call was tracked with the source and outcome so at any point in the last 30 days they know exactly which calls came from which channel and which ones booked and that’s the baseline you need before you can evaluate a channel like ChatGPT. And if you don’t have that baseline, you don’t know what’s working and you definitely won’t know when ChatGP starts sending you leads. How the bear dashboard works.

Stats infographic showing ChatGPT call conversion and lead volume data
Bear Restoration logged 82 leads in 30 days, with every phone call tracked by source and outcome.

In the bear dashboard, calls sit right next to forms and LSA leads. So you’re not toggling between three different platforms. You’re looking at one screen that shows where the lead came from, when it came in, and what happened to it.

And when a call comes in as a tagged AI referral, it doesn’t get lost in the spreadsheet it shows up in context next to everything else that matters because AI referred call volume is low right now so if you’re only checking once a month you’ll miss two or three calls that actually came through you’ll need to see them in real time next to your other channels so you can judge whether the call volume is trending up the tracking setup you need here’s the minimum setup I’d recommend one track phone number for your website one for your Google business profile and one for your paid search if you’re running it, and a system that asks the source question at every inbound call.

In this video

  1. 0:00 YouTube Videos Driving AI Citations
  2. 0:24 Who Jeremy Is and Why This Matters
  3. 0:40 Four Weeks of PushLeads Citation Data
  4. 1:20 Citation Breakdown by Destination Type
  5. 1:50 AI Visibility Score Video vs Website
  6. 2:09 Google AI Overview Keyword Findings
  7. 2:40 Google vs ChatGPT Claude Perplexity
  8. 3:19 Strategy by Engine and Content Type
  9. 3:43 Thin Watch Pages Almost Never Cited
  10. 4:06 Embedded Video Pages Citation Edge
  11. 4:41 Tracker Gap Understating Your Numbers
  12. 5:11 Six Actions We Are Taking Now

Full transcript: Do YouTube Videos Help AI Visibility? 4 Weeks of Real Data

This doesn’t have to be complicated, and a lot of call tracking platforms will let you record the intake call and log the source automatically. And if you’re a Pushleads client, this is already built into our own dashboard. If you’re not, the principle is the same. Every lead source gets its own phone number, and every call gets logged, and you review it weekly. What ChatGPT actually sees when you look it up. Here’s something that most business owners don’t know. ChatGPT doesn’t just search the web the same way that Google does. It’s pulling from the mix of its training data, real-time web browsing in some cases, and third-party data sources like Yelp, Angie, and other directories.

So if your business information is inconsistent across all these platforms, ChatGPT may not recommend you or may recommend you with the the wrong details. Your name, address, and phone number all need to match up everywhere. And this is called the NAP. And this is also important for the Google Business Profile as well. So your reviews also matter on these platforms. The way people describe your services on review sites is part of what ChatGP reads when it decides who to recommend. What gets you named in ChatGPT? In my experience working on AI visibility, the businesses that get named most reliably have a few things in common. Consistent information everywhere.

Want more customers from Google & AI search?

Get a free SEO audit of your site — see exactly what to fix first.

Get My Free Audit Book a Call

Up next

The AI Search Revolution: How 2 Billion People Are Finding Information in 2025

11 min read · AI search is reshaping how billions find information in 2025, fundamentally transforming the marketing landscape and requiring immediate strategy shifts.

Want this done for you?

Scan your site. See the gaps. Decide with numbers.

  • Rankings, Google Business Profile and site-speed teardown
  • Report lands in your inbox in about a minute
  • No obligation · 4.9 from 67 Google reviews

Prefer a human? Book 20 minutes with Jeremy

Continue reading