What actually makes ChatGPT or Google AI recommend one local business instead of another?

Andrew Shotland, founder and CEO of Local SEO Guide, joins Jason AI Wade to break down what his team is testing right now across local SEO, AI visibility, citations, service pages, and multi-location businesses.

Andrew’s core point is simple: the fundamentals of SEO still matter. A machine has to understand what your business does before it can recommend you. The difference now is that AI systems are assembling that understanding from more pieces of evidence and using it inside conversational answers.

That changes how businesses should think about visibility.

It is not enough for a plumber to say, “We’re a plumber.”

If the business wants to be recommended for water heater replacement, emergency service, or storm drains, those specific services need to appear clearly across the company’s website and the third-party sources AI may use to corroborate them.

Andrew describes those external citations almost like additional sides of a die: the more relevant sources that connect a business with the right location, service, and attributes, the more opportunities an AI system has to encounter the same association.

Then he shares a real test.

His team analyzed search data across a 160-location company and found a missing concept around construction job-site bathrooms and toilets. They added that language to the Google Business Profiles and local citation network for 20 locations.

Within days, visibility changed in Google AI Mode.

Later, the test locations began improving in ChatGPT while the untouched locations initially remained flat.

Roughly a month later, visibility began increasing across the wider network as well.

Andrew’s working theory: enough pages had connected the brand with the concept that AI systems began applying that relationship more broadly across the entity.

The conversation also gets into competitor-focused content, local search strategy, agentic SEO, automation, content-gap analysis, and what Andrew calls the “infinite conversational paths” created by AI search.

The practical lesson is not to wait for a perfect GEO playbook.

Test something.

Measure whether visibility changes.

Then apply what works.

Andrew compares the process to baseball: even a great hitter misses most of the time. The advantage comes from taking enough intelligent swings.

Andrew Shotland
Founder and CEO of Local SEO Guide.

Andrew has worked in search for roughly two decades and now focuses heavily on local SEO, AEO, multi-location businesses, and how AI systems understand and recommend companies.

He describes the current period as one of the most interesting moments he has experienced in search.

AI recommendations for local businesses
Local SEO and GEO
Google AI Mode
ChatGPT visibility
Local citations
Entity corroboration
Service-page strategy
Content gaps
Competitor alternatives
Conversational search
Agentic SEO
Multi-location testing
AI visibility experiments

Andrew Shotland / Local SEO Guide
https://www.localseoguide.com/

Andrew Shotland on LinkedIn
Andrew identifies LinkedIn as his main social platform and where he posts most often.

AI Visibility Podcast
BackTier

Why Consistent Anchor Information Matters in AI Visibility

One of the most overlooked problems in AI visibility is consistency.

Companies tend to think about their website, LinkedIn page, directory listings, biographies, press mentions, podcasts, and third-party profiles as separate marketing assets. AI systems do not necessarily experience them that way. They encounter all of these sources as evidence about the same entity.

That creates a simple but important problem: if the basic facts keep changing depending on where the system looks, confidence in the entity can weaken.

I call these basic facts anchor information.

Anchor information is the relatively stable information that helps a machine determine who or what an entity is. For a company, that includes the company name, founders, location, primary category, core services, major products, important dates, leadership, and the language used to describe what the business actually does.

For a person, it may include name, professional role, company relationships, areas of expertise, location at an appropriate level, published work, and other persistent associations.

None of this sounds particularly sophisticated. That is exactly why companies underestimate it.

The problem begins when the same organization describes itself five different ways across five different sources. One profile says the company is an SEO agency. Another describes it as an AI consultancy. The website calls it an AI visibility company. An older directory calls it a digital marketing agency. The founder's biography uses yet another category.

A human can usually reconcile those descriptions.

A machine has to decide whether they represent complementary facts, historical changes, marketing language, or genuinely different classifications.

That uncertainty becomes more important as AI systems move beyond finding pages and begin comparing entities, selecting sources, recommending companies, and answering questions directly.

The objective is not to make every page identical. That would be unnatural and unnecessary.

It is an information architecture problem.

Before an AI system can cite a business, recommend it, or associate it with a particular area of expertise, the system first has to determine what the entity is and what facts about it can be trusted.

That starts with anchors.

A company should be able to answer a basic question: if an AI system encountered our website, founder biography, LinkedIn profile, podcast appearances, business directories, press coverage, and third-party references, would those sources collectively describe the same organization?

His central thesis is straightforward: before a company can influence how AI describes it, the company first has to make the truth about itself clear enough for machines to understand.

Jason Wade
https://jasonwade.com

BackTier
https://backtier.com

NinjaAI
https://ninjaai.com

The Democratization of AI Agents

When Everyone Can Build a Bot That Actually Does Something

For years, building software that could automate business processes required programmers, expensive integrations, and technical expertise.

Now AI agents are making automation accessible to people who have never written a line of code.

A small business owner can describe a task in plain English and increasingly use AI tools to build an agent that researches information, handles customer inquiries, updates records, or coordinates work across multiple systems.

What once required a development team may now be possible with a subscription and some experimentation.

The next AI revolution isn't just about everyone having access to intelligence. It's about everyone having access to automated labor.

• From Chatbots to Agents: Why answering questions and taking actions are fundamentally different capabilities.

• The No-Code Revolution: How natural-language tools are lowering the technical barriers to building AI agents.

• Bots for Everyone: Why small businesses, independent professionals, and entrepreneurs can now experiment with automation previously reserved for larger organizations.

• The One-Person Company: How agents can help individuals operate workflows that once required multiple employees or contractors.

• Agents Working Together: How specialized agents can delegate tasks, exchange information, and coordinate across business systems.

• The Agent Marketplace: What happens when businesses can buy, customize, or deploy agents instead of developing every automation themselves.

• The Reliability Problem: Why agents can still make mistakes, misunderstand instructions, and fail during multistep tasks.

• The Security Question: What happens when an inexpensive bot has access to customer records, financial systems, or company communications?

• The Agent Economy: How automated software may change service businesses, employment, pricing, and competition.

• The AI Visibility Connection: Why companies may eventually need to serve both human customers and the AI agents acting on their behalf.

The internet democratized publishing.

Cloud software made powerful business applications accessible through subscriptions.

Generative AI made sophisticated writing, coding, and analysis available through conversational interfaces.

AI agents extend that progression from generating information to executing work.

But access doesn't guarantee competence.

An agent that can operate software still needs accurate instructions, appropriate permissions, reliable tools, monitoring, and clear boundaries.

The businesses that benefit most may not be those deploying the greatest number of agents, but those identifying valuable workflows and making automation dependable.

AI is lowering the barrier between having an idea for automation and actually deploying it.

That creates opportunities for small businesses and independent operators, but it also introduces new risks when automated systems act without sufficient oversight.

The competitive advantage will increasingly come from knowing what to automate, how to connect systems, and when human judgment is still required.

The internet gave everyone a voice. AI agents are beginning to give everyone a workforce.

Presented by BackTier | NinjaAI.com

Topics: AI Agents, Democratization of AI, Agentic AI, No-Code Automation, AI Bots, Autonomous Agents, Small Business AI, AI Workforce, Multi-Agent Systems, AI Economy, AI Visibility.

What We'll CoverThe Bigger PictureKey Takeaway