Let's Talk About ChatGPT Replacing Word-of-Mouth for Local Businesses

Discussion about how ChatGPT recommendations drive real traffic and what it means for local service businesses trying to stay visible in the AI era.

Read the full blog post: https://go.speedmobi.com/chatgpt-recommendations-drive-traffic-pod

Transcript

SPEAKER_02 0:00

Imagine a trusted friend, right, whispering a business recommendation into like millions of ears at the exact same moment.

SPEAKER_00 0:07

Aaron Powell Which would be the dream scenario for any business.

SPEAKER_02 0:11

Right, exactly. You'd expect a massive spike in sales and you know a giant flashing neon sign pointing exactly to where all those new customers came from.

SPEAKER_00 0:18

Oh, absolutely. You'd want to track every single one of them.

SPEAKER_02 0:21

Aaron Powell But um what if your tracking software couldn't see a single one of them?

SPEAKER_00 0:24

Aaron Powell Yeah. That's where it gets complicated.

SPEAKER_02 0:27

Welcome to this deep dive. Today we are unpacking how AI chat bots have secretly become like the most powerful, totally hidden traffic drivers for local businesses.

SPEAKER_00 0:39

Aaron Powell It's a huge blind spot for so many companies right now.

SPEAKER_02 0:41

Aaron Powell It really is. And our mission today is to uncover this invisible traffic multiplier. We're going to figure out why traditional analytics are just completely blind to it. Right. And then we'll break down the actual mechanics of how these new AI algorithms choose who actually gets recommended.

SPEAKER_00 0:57

Aaron Powell And it is a crucial topic, especially right now. I mean, we are looking at the absolute earliest stage of a massive shift in customer discovery. This isn't just like a cool new feature on the internet. It's a fundamental rewiring of how consumers find services and honestly, how businesses are either found or totally forgotten.

SPEAKER_02 1:17

Aaron Powell And to guide us through this, we're pulling from a massive stack of research today. We've got a really comprehensive news study from SimilarWeb that just dropped in June 2026.

SPEAKER_00 1:28

Incredible data in that one.

SPEAKER_02 1:29

Oh yeah. Plus, we are looking at the prompt-to-purchase research conducted by Michael Yanelli and Alan I.

SPEAKER_00 1:36

Right, which really nails down the behavioral side of things.

SPEAKER_02 1:39

Exactly. And we're tying it all together with the latest analysis from the SpeedMoby research team. So, you know, if you're a business owner or maybe you're prepping for a strategy meeting or just intensely curious about the mechanics of our new internet, you're going to want to hear this.

SPEAKER_00 1:53

You definitely are.

SPEAKER_02 1:54

Okay, let's unpack this. We need to start with the sheer scale of what these AI recommendations are actually doing to web traffic.

SPEAKER_00 2:02

Yeah, because to really understand the shift, you have to look at the explosive numbers. I mean, the Similar Web study provides the first really large-scale quantitative evidence that AI-powered recommendations are moving way past the novelty phase.

SPEAKER_02 2:17

Right. It's not just people asking it to write poems anymore.

SPEAKER_00 2:19

Exactly.

SPEAKER_02 2:20

Yeah.

SPEAKER_00 2:20

They're creating measurable brand discovery behavior and driving hard actual website traffic.

SPEAKER_02 2:26

Aaron Powell And according to their data, brands that get recommended by Chat GPT receive two and a half times more website visits within a seven-day window compared to brands that aren't recommended. A 2.5x traffic multiplier. That is huge.

SPEAKER_00 2:41

It's massive. And this was a totally consistent effect across the major industries they studied too. Trevor Burrus, Jr.

SPEAKER_02 2:46

Specifically finance, travel, and beauty, right?

SPEAKER_00 2:48

Trevor Burrus Right. They focused this particular slice of research on US desktop activity. But they're already planning to expand it. So the volume is very clearly there.

SPEAKER_02 2:56

Aaron Powell But you know what what really caught my eye were the engagement metrics, the quality of this traffic. Trevor Burrus, Jr.

SPEAKER_00 3:01

Oh, the quality of that traffic is what fundamentally changes the game.

SPEAKER_02 3:04

Let's talk about the pages per session, because that blew my mind. Aaron Powell Yeah.

SPEAKER_00 3:07

So standard internet visitors, according to the similar web baseline, they view an average of 6.5 pages per session.

SPEAKER_02 3:14

Okay.

SPEAKER_00 3:15

But the AI referred visitors, they are viewing 12 pages per session.

SPEAKER_02 3:19

Wait, hold on. 12 pages a session?

SPEAKER_00 3:21

Yes. That is an 84.6% increase.

SPEAKER_02 3:24

But are we are we absolutely sure those aren't just like AI bots or web scrapers crawling the site to gather data? I mean, how do we know these are actual human buyers clicking around that much?

SPEAKER_00 3:34

Aaron Powell That's a great question, but the time on site data confirms these are humans doing actual deep research.

SPEAKER_02 3:40

Okay. How long are they staying?

SPEAKER_00 3:42

Well, a standard visitor stays for about 5.6 minutes.

SPEAKER_02 3:45

Which is already pretty decent for local business sites.

SPEAKER_00 3:49

All right. But the AI referred visitor stays for 11.8 minutes.

SPEAKER_02 3:52

Whoa.

SPEAKER_00 3:53

Yeah, that's 110.7% increase. Bought scrape a site in seconds. You know, they don't hang around. These users are spending nearly 12 minutes reading. Wow. You really have to look at the psychology of the user journey to understand why that's happening. Why does someone open an AI like ChatGPT instead of just, you know, browsing social media or doing a quick search? It's because they have a specific immediate problem to solve.

SPEAKER_02 4:14

So they aren't window shopping. To use an analogy, this is the difference between a casual browser and a buyer walking in with a clipboard and a mission.

SPEAKER_00 4:22

Exactly. That's a perfect way to put it. When these AI referred visitors hit a website, they are already in an active consideration mode. Oh they're evaluating services, comparing pricing, really reading the fine print. The AI has essentially done the top of funnel work by pre-qualifying the business for them.

SPEAKER_02 4:40

Okay, so the trust is already established before they even click.

SPEAKER_00 4:43

Exactly. So when they finally arrive on the site, they are highly motivated, pre-warmed leads. They are totally ready to do the deep work of making a purchasing decision.

SPEAKER_02 4:52

Okay, but this raises a huge question for me. If this traffic is essentially doubling their time on site, viewing twice as many pages and arriving totally ready to buy, why aren't marketing agencies bragging about this from the rooftops? I mean, if I'm a local business owner and my highly qualified traffic just 2.5x, I'd want to know exactly where it came from so I could like pour more money into it.

SPEAKER_00 5:17

And that is the problem. Because the businesses and the agencies, they can't see it.

SPEAKER_02 5:23

Wait, they can't see it at all.

SPEAKER_00 5:25

No, this is the ghost traffic problem. And this transition from the immense value of this traffic to its total invisibility is the crux of why this technology is so incredibly disruptive.

SPEAKER_02 5:36

Let's break down this attribution challenge because it feels bizarre that in 2026, with all of our hyper-advanced tracking software, we are essentially flying blind.

SPEAKER_00 5:44

We really are.

SPEAKER_02 5:45

The similar web data points out that 55.9% of this AI-driven traffic arrives through branded Google searches. Meaning the AI chat butt doesn't usually just hand the user a convenient clickable link.

SPEAKER_00 5:57

Right. Think about the mechanics of a conversational AI. You ask it for the um the best boutique hotels in Austin, Texas.

SPEAKER_02 6:04

Okay.

SPEAKER_00 6:05

It synthesizes an answer and mentions a specific hotel by name in a totally natural sentence. It often doesn't provide a hyperlink at all.

SPEAKER_02 6:12

Right.

SPEAKER_00 6:14

Exactly. So what do you, the user, do next?

SPEAKER_02 6:17

I'd probably open a new tab.

SPEAKER_00 6:18

Right. You open a new tab, go to Google, and type in the exact name of that specific hotel the AI just told you about.

SPEAKER_02 6:26

And when you do that, the hotel's Google Analytics dashboard records your arrival as an organic search.

SPEAKER_00 6:33

Yep.

SPEAKER_02 6:33

Or I guess if you type the URL directly into the bar, it shows up as a direct visit. There is no UTM parameter. You know those little tracking codes you see tagged at the end of a web address that tell a business exactly what ad you click.

SPEAKER_00 6:45

Exactly. The tracking chain is totally broken. And last click attribution, you know, the software model that gives 100% of the credit to the absolute last thing a user click before buying.

SPEAKER_02 6:57

Right, which everyone uses.

SPEAKER_00 6:59

Which everyone uses. It completely misses this. The business thinks their organic SEO is doing great, totally unaware that ChatGPT actually generated the lead. That's wild. Or worse, they're losing leads to an AI recommended competitor without ever knowing they were even in the running in the first place.

SPEAKER_02 7:15

Wow. And this invisible pipeline is confirmed in the prompt-to-purchase study, right? The one by Michael Ianelli and Alan I.

SPEAKER_00 7:22

Yes, absolutely.

SPEAKER_02 7:23

Because they looked at what happens when an AI, whether it's ChatGPT, Claude, or Gemini, recommends a brand to users who had zero measurable prior engagement with that company.

SPEAKER_00 7:36

And the results there are the absolute smoke and gun.

SPEAKER_02 7:38

Yeah.

SPEAKER_00 7:39

For users with zero prior engagement, an AI recommendation caused a 4.3 percentage point jump in same-name Google searches.

SPEAKER_02 7:47

Okay, so that proves the search behavior.

SPEAKER_00 7:50

Right. It also drove a 2.4 percentage point rise in direct website visits and a 1.2 percentage point growth in brand-specific retailer page visits.

SPEAKER_02 7:59

Aaron Powell Here's where it gets really interesting to me. So the AI is literally creating this search behavior out of thin air. It mentions the name, and the human independently verifies it on Google.

SPEAKER_00 8:09

Exactly.

SPEAKER_02 8:10

It's not a traditional digital click path. It's like a whisper network.

SPEAKER_00 8:14

Aaron Powell That's a great way to describe it. It mirrors the exact mechanics of a real-world personal referral. Right. If I tell you about a fantastic new coffee shop, you don't magically teleport there, and I don't hand you a URL.

SPEAKER_02 8:25

No, I pull out my phone.

SPEAKER_00 8:26

You pull out your phone, you Google the name I just said, check their hours, and then you go.

SPEAKER_02 8:29

And the coffee shop's tracking software just sees a Google search.

SPEAKER_00 8:33

Exactly. It has no idea I was the catalyst. AI functions as a digital word-of-mouth engine.

SPEAKER_02 8:38

But the monumental difference is scale.

SPEAKER_00 8:40

Right. I can tell two friends, and AI can make millions of these hyper-personalized recommendations simultaneously.

SPEAKER_02 8:46

Okay. So we know users are Googling the brand names after the AI recommends them. Yep. But what happens to the users who bypass the chat bot entirely and just search for a generic service directly on Google? Because Google itself has radically changed, right?

SPEAKER_00 9:01

Oh, completely.

SPEAKER_02 9:02

This connects the Chat GPT multiplier to what industry analysts are calling the zero-click reality.

SPEAKER_00 9:08

And this is where the landscape gets genuinely existential for local businesses. Let's look at the sheer scale of AI search usage today. At Google I.O. in May 2026, they confirmed that over one billion people are using AI mode in Google search every single month.

SPEAKER_01 9:26

One billion.

SPEAKER_00 9:27

One billion. And those AI search queries are doubling every quarter.

SPEAKER_02 9:31

That is staggering. And the ecosystem is heavily monetizing too. I mean, ChatGPT has globalized its advertising platform to more than eight countries now, with over 1,400 brands paying for recommendations right now. But let's look at the SparkToro and similar web data regarding zero-click searches. Because if you rely on traditional SEO, these numbers are terrifying.

SPEAKER_00 9:52

They really are.

SPEAKER_02 9:53

According to their research, 68.01% of US Google searches in early 2026 ended without a single click. Wow. More than two-thirds of the time someone searches for something, they never click a link to visit a website. And that is up 7.56 points from just 2024.

SPEAKER_00 10:10

Right. And when you factor in AI overviews specifically, the behavior is even more extreme. AI overviews trigger an 83% zero click rate.

SPEAKER_02 10:18

Okay, so so what does this all mean for a local business? If I'm a local plumber listening to this, are you saying that trying to rank as the number one blue link on Google is basically a dying strategy? Honestly, yes.

SPEAKER_00 10:29

Wow. The paradigm has fundamentally shifted. For two decades, the goal of a business was to appear in the blue links below the answer.

SPEAKER_02 10:36

Right. You wanted the user to click to your site to get the answer.

SPEAKER_00 10:39

You want it to be the final destination. But in a world with an 83% zero click rate on AI overviews, the AI is the destination.

SPEAKER_02 10:48

Wow.

SPEAKER_00 10:49

The AI reads the internet, synthesizes the information, and gives the answer directly to the user right there on the search page. You can no longer aim to appear below the answer.

SPEAKER_02 10:59

You have to be in the answer.

SPEAKER_00 11:00

You must now appear inside the AI's synthesized answer itself. Trevor Burrus, Jr.

SPEAKER_02 11:04

Because if the AI doesn't choose to include you in its summary, you just don't exist to that user.

SPEAKER_00 11:11

Precisely. You are entirely invisible. And as we established with the ghost traffic data, you are bleeding those highly qualified clipboard holding leads to your competitors without your analytics dashboard ever showing a dip.

SPEAKER_02 11:22

Aaron Powell That is such a stark reality. If the old playbook of fighting for the top blue link is burning, we clearly need a new playbook.

SPEAKER_00 11:29

We do.

SPEAKER_02 11:29

We know the opportunity is massive, and we know the zero-click risk is real. So let's shift to the mechanics. How do we actually get the AI to pick us?

SPEAKER_00 11:38

Let's get into it.

SPEAKER_02 11:39

If I'm that local plumber, how do I get inside the AI's synthesized answer?

SPEAKER_00 11:44

Well, the Speed Moby research team and current industry analysis have basically reverse-engineered what these large language models are actually looking for when they decide who to recommend.

SPEAKER_01 11:54

Okay.

SPEAKER_00 11:54

And it starts with how the AI understands your own website's content.

SPEAKER_02 11:58

Right. But writing a clear website makes sense on a basic level. I mean, anyone can hire a copywriter to write a good paragraph about fixing pipes. Sure. How does an LLM process that differently than the old Google crawler that just looked for the keyword, you know, Austin Plumber 20 times on a page?

SPEAKER_00 12:15

Aaron Powell Because LLMs are semantic engines, not keyword matchers.

SPEAKER_02 12:18

Semantic engines.

SPEAKER_00 12:19

Right. They are trained on natural human language, so they prioritize natural language. When the AI scans your site, it isn't counting keywords anymore. It's trying to understand the context of what you actually do. This means you need highly detailed descriptions of your services written in plain English.

SPEAKER_02 12:38

Not just robot SEO speak.

SPEAKER_00 12:40

Exactly. And you need geographic specificity. Not just saying Austin, but mentioning specific service areas and local neighborhoods.

SPEAKER_02 12:47

That makes sense.

SPEAKER_00 12:48

And crucially, your website needs to feature FAQs that mirror how real living people actually ask questions.

SPEAKER_01 12:56

Give me an example.

SPEAKER_00 12:57

Like instead of plumbing services, Austin, an FAQ should be how much does it cost to fix a leaky pipe in South Austin? Because if your website already answers a conversational question natively, the AI will confidently pull from it to answer the user's conversational prompt.

SPEAKER_02 13:13

Okay, so clear. Natural language is the baseline. It sounds like you can't just have a flashy storefront anymore. You have to feed the AI a perfectly organized filing cabinet instead of a messy desk, or it won't know how to categorize you.

SPEAKER_00 13:24

That is a brilliant analogy.

SPEAKER_02 13:26

But again, anyone can write good copy about themselves. How does the AI know I'm actually good at what I do? It can't just take my word for it, right?

SPEAKER_00 13:33

It doesn't. Which brings us to how the AI verifies your claims using review signals.

SPEAKER_02 13:38

Okay.

SPEAKER_00 13:39

But there is a massive shift here based on Google's March 2026 policy update.

SPEAKER_02 13:45

Oh, right. What did that change?

SPEAKER_00 13:46

Aaron Powell Well, it used to be about sheer volume, right? Whoever had the most five-star clicks won. Now the AI algorithms heavily weight review quality over volume.

SPEAKER_02 13:58

Because the AI is actually reading the text of the reviews to build semantic context. Meaning a detailed three-paragraph review about how a plumber fixed a specific, highly unusual vintage sink is worth more to the AI than like 20 reviews that just say five stars, great job.

SPEAKER_00 14:16

Exactly. The AI needs contextual depth to feel confident in making a recommendation.

SPEAKER_01 14:20

That's amazing.

SPEAKER_00 14:21

It looks for review diversity too. Like do the reviews mention different specific services? It looks at recency and it doesn't just look at Google anymore.

SPEAKER_02 14:29

Where else is it looking?

SPEAKER_00 14:30

Cross-platform presence is vital now. The AI is scanning Google, Yelp, and industry-specific directories to cross-reference your reputation. It's looking for consensus across the entire web.

SPEAKER_02 14:41

Okay, so the AI reads my site, understands what I claim to do, and then reads my reviews to verify that I actually do it well.

SPEAKER_00 14:47

Right.

SPEAKER_02 14:48

But what about the rest of the web? Does the AI care if anyone else is talking about me?

SPEAKER_00 14:52

Yes, very much so. Through third-party mentions and citations.

SPEAKER_02 14:56

Okay.

SPEAKER_00 14:57

The AI wants proof that you exist as a relevant entity in the real world. This means local business directories, features in local news, or mentions on industry-specific blogs.

SPEAKER_02 15:08

So not just your own properties.

SPEAKER_00 15:10

Exactly. If authoritative third-party sites are talking about your business, it acts as a massive trust signal for the LLM.

SPEAKER_02 15:17

Aaron Powell Which leads to a concept that sounds incredibly basic, but is apparently the place where most businesses just completely fail.

SPEAKER_00 15:24

Oh, I know what you're going to say.

SPEAKER_02 15:25

And AP consistency. Name, address, and phone number.

SPEAKER_00 15:28

Yes.

SPEAKER_02 15:29

I like to think of the AI as a really incredibly smart but blindfolded librarian.

SPEAKER_00 15:34

Oh, I like that.

SPEAKER_02 15:35

Right. It can't look at your flashy physical storefront. It can only feel the shape of the data blocks you hand it.

SPEAKER_00 15:41

Exactly.

SPEAKER_02 15:41

If the address data block on your Yelp page doesn't perfectly match the exact shape of the address block on your website, it just drops it.

SPEAKER_00 15:48

That is exactly how it works. Yeah. The AI is a hyperlogical machine that absolutely hates risk. Right. If your Google Business Profile says you are open until 6 PM, but your website says 5 p.m. and your Yelp page says you're closed on Tuesdays.

SPEAKER_02 16:03

The AI freaks out.

SPEAKER_00 16:04

The AI loses all confidence. It will not risk its own reputation by recommending a business with conflicting data to a user. You must have absolutely matching service descriptions, hours, and contact info across every directory on the internet.

SPEAKER_01 16:20

It's fascinating because none of this requires a massive advertising budget right now.

SPEAKER_00 16:23

No, it doesn't. If we connect this to the bigger picture, optimizing for AI right now is exactly like the early days of local SEO back in the late 2000s.

SPEAKER_02 16:32

So wow. Yeah.

SPEAKER_00 16:33

You don't need to outspend your competitor. You just need to outorganize them.

SPEAKER_02 16:36

That's a great takeaway.

SPEAKER_00 16:38

It requires clear, natural content, deep contextual reviews, and absolute data consistency. And the best part is the compounding effect.

SPEAKER_02 16:46

Right. Because if you organize your data perfectly, the AI recommends you.

SPEAKER_00 16:50

Yep.

SPEAKER_02 16:50

The user then Googles your exact brand name, which creates that ghost traffic we talked about. And then Google's algorithm sees that sudden surge in people searching specifically for you and says, wow, people really want this business, which in turn drives your traditional organic rankings even higher.

SPEAKER_00 17:07

It is a profound self-feeding cycle. AI visibility drives branded search, which then cements long-term SEO dominance.

SPEAKER_02 17:17

We have covered a staggering amount of ground today.

SPEAKER_00 17:20

We really have.

SPEAKER_02 17:20

I mean, we started with the sheer volume of a 2.5x traffic multiplier. We waded through the attribution nightmare of ghost traffic.

SPEAKER_00 17:28

Which is terrifying.

SPEAKER_02 17:29

Truly. We confronted the 83% zero-click reality, and we broke down how an LLM actually verifies data to make a recommendation. The full picture. So the takeaway here for you listening is incredibly clear. The window to establish dominance in this new AI discovery channel is wide open right now, but it is going to close fast.

SPEAKER_00 17:47

The first mover advantage cannot be overstated here. As more businesses realize their traditional analytics are basically lying to them by a mission, the competition for AI inclusion is going to skyrocket.

SPEAKER_02 17:58

Before we go, I want to quickly synthesize the immediate action steps from the SpeedMoby research team because you can actually start testing this today.

SPEAKER_00 18:05

Oh, definitely this.

SPEAKER_02 18:06

First, test your own visibility right now. Open ChatGPT or Google AI mode and search for your service and your location. See if you even exist to the algorithm. That's step one. Second, audit your website's copy. Ensure you are using natural language and hyper-specific geographic markers. Third, stop asking for generic five-star ratings and start encouraging your customers to write detailed service-specific reviews.

SPEAKER_00 18:33

Yes, give the AI context.

SPEAKER_02 18:35

Exactly. Fourth, do a complete audit of your name, address, and phone number across every single directory online. Make sure that blindfolded librarian fills the exact same data blocks everywhere.

SPEAKER_00 18:47

Perfect consistency.

SPEAKER_02 18:48

Fifth, cultivate those third-party mentions. And finally, monitor your branded search volume in your analytics. Since you can't track the AI directly, a spike in people searching your exact name is your best proxy that the AI is working for you.

SPEAKER_00 19:02

Those are the practical, foundational steps every business needs to take immediately. Absolutely. But you know, as we step back and look at this massive technological shift, it raises a critical, almost philosophical question about the future of commerce for me.

SPEAKER_02 19:15

Oh, what's that?

SPEAKER_00 19:17

Well, if AI acts as the ultimate hyperlogical gatekeeper for this new digital word of mouth, and if it strictly requires a perfect, deep digital paper trail of contextual reviews, local news mentions, and flawless directory consistencies just to feel confident enough to recommend you, what happens to the brand new, wildly innovative local businesses?

SPEAKER_01 19:39

Oh, wow.

SPEAKER_00 19:40

Right? Are we moving toward a web where only the oldest, most documented companies can survive the algorithm while the new innovators remain entirely invisible?

SPEAKER_02 19:50

That is definitely something to think about. Thank you so much for joining us on this deep dive. Keep questioning, keep learning, and we'll see you next time.