A website needs an outcome
Cory and Wences have both seen the same challenge play out over and over.
A business gets a new website. It looks good, has the right information, and does what it was built to do, but then it sits there.
Wences sees this pattern at scale. Through marketgoo's partnerships, his team has worked with millions of websites, many of which are five to seven pages and function primarily as online portfolios.
Meanwhile, discovery is becoming more fragmented.
Wences commented:
The problem has been the same in the past, but today the problem is a little bit more sophisticated. Discovery is fragmented.
Potential customers might find a business through Google Search, Maps, reviews, an AI assistant, or another source entirely. Keeping up with all those channels can quickly become another job for a small business owner.
That's why Wences starts with a different question: What does the business want to accomplish?
A pizza shop wants more people through the door. A consultant wants more qualified inquiries. A local service business wants nearby customers to find them.
As Wences put it:
I don't want better SEO. I want more visits.
That thinking is shaping marketgoo's work on Flavio, an AI agent designed for very small businesses. Flavio is being built to understand the business and its goals, then take on more of the work required to improve its online visibility.
Wences describes it as a "growth layer" that comes after a business establishes its online presence.
The starting point is context: what the business does, where it operates, what customers are saying, and what it wants to achieve. From there, the system can determine which actions support those goals.
Context gives AI something useful to work with
AI has made it easy to generate content, answer questions, and automate individual marketing tasks. But giving a business owner another blank prompt box doesn't necessarily solve their problem.
They still need to know what to ask.
Wences sees that as a poor fit for many of the small businesses marketgoo serves.
As Wences put it:
If I'm a business owner and you give me all the ability, still I need to have the time or understand which questions can I prompt?
Flavio takes a different approach. The goal is for the agent to take the lead once it understands the business and what the owner wants to accomplish.
That could include improving a Google Business Profile, identifying service areas, helping the business appear in more relevant local searches, analyzing reviews, or improving website content.
One example Wences shared involves FAQs.
The agent can look at information already available on a website, combine that with reviews and other context about the business, and identify questions potential customers may be asking. It can then help add useful answers to the website, making that information easier for people and AI systems to understand.
The business owner doesn't need to know that an FAQ could help with search visibility or AI discovery. They don't need to determine which questions to target.
Their expertise is their business. The technology handles more of the marketing mechanics.
AI agents need independence and guardrails
Giving AI more responsibility raises another question: how much should an autonomous system be allowed to do?
Wences doesn't envision an agent constantly asking the business owner what to do next. That would simply create a different kind of work.
He compared it to having someone on your team:
You don't want that person to ask you all the time, 'Okay, what should I do? What should I do?'
The goal is to establish enough context, direction, and trust that the agent can work independently within defined boundaries.
That doesn't mean giving AI unlimited control.
Wences explained that Flavio combines AI with deterministic systems, defined focus areas, and guardrails around what the agent should and shouldn't do. Users will also be able to review and roll back implementations.
Some actions may require approval. Others can happen in the background.
For agencies considering how AI agents might fit into client work, that balance matters. Automation becomes more useful when it reduces work without removing appropriate oversight.
Search is becoming less predictable
The conversation eventually turned to a question almost everyone working online is asking: What happens to SEO now?
Wences has spent much of his career in SEO, but even he isn't pretending there's a settled answer.
What is clear is that search behavior is changing.
Businesses that once relied heavily on organic search are seeing more zero-click experiences, where users get an answer without visiting the source website. AI tools add another discovery layer, with responses assembled from websites, reviews, communities, videos, and other sources.
The result is a marketing environment that Wences describes as more probabilistic.
The old model could be relatively direct: identify keywords, create and optimize content, monitor rankings, and iterate.
AI discovery introduces far more context. The answer one person receives can depend on the exact question they're asking, their needs, and the information available to the model.
Through all of that change, Wences sees familiar fundamentals holding up.
Create good content for the people you want to reach. Structure it well. Make sure your site performs well. Give visitors a good experience. Make it easy for people to understand who you are and what you offer.
Those fundamentals still matter. In fact, many of the practices that help people find and understand your content can also help AI systems do the same. We explored that idea in more detail in our livestream, SEO for LLMs: It's just SEO (mostly).
Your brand may matter more than another optimization trick
As search and AI discovery become harder to predict, Wences believes businesses should pay more attention to something they can control: their brand.
The better marketing is a good brand.
That means understanding what makes your business useful and different, then making sure the rest of the world can understand it too.
For local businesses, that story exists across more than the website.
It's in customer reviews. It's in Google Business Profile. It's in how services are described. It's in the questions customers ask and the answers a business provides. It's in the places where people talk about and recommend the company.
Those signals can help people build trust with a business. They can also give search engines and AI systems more context about when that business might be relevant.
The technology around discovery may change. Being clear, useful, trustworthy, and easy to understand remains a strong foundation.
What does this mean for agencies?
Flavio is initially focused on micro and small businesses, but Wences sees an interesting opportunity for agencies too.
Many agencies already face the problem Cory described at the beginning of the conversation. They can build a great website and keep it running, but the client may need ongoing marketing help that falls outside the agency's expertise, budget, or service model.
AI agents could help fill some of that space.
Wences described conversations with agencies interested in deploying agents across client websites. Instead of handing a client another marketing tool to learn, an agency could potentially provide an agent that continues working on the client's online presence after the site launches.
You're an agency and you have some customers, and you want to delegate all of the marketing. You want to make them successful online, but you don't have the ability, or they don't want to hire anybody, or it's too complex.
Wences is quick to point out that this model is still early. Marketgoo plans to work closely with users first, learn how Flavio performs in the real world, and explore an agency layer as the product develops.
But the idea opens an interesting door.
The end of a website project doesn't have to be the end of the work that website does for the business.
Focus on what the business is trying to accomplish
SEO isn't disappearing overnight. Neither is Google. AI isn't giving us a new marketing playbook that replaces everything that came before it.
What we're seeing is a wider set of ways for people to discover businesses. That makes it even more useful to step back from individual channels and focus on what you're trying to accomplish.
For a small business owner, that might mean getting more people through the door, generating more inquiries, or reaching customers who need a particular service.
Start with that outcome. Then think about where those customers are looking, what they need to know, and how your online presence can help them get there.
For agencies, the same questions can help shape what happens after a site launches.
AI agents like Flavio offer one possible way to close the gap between having an online presence and putting that presence to work. As these tools develop, the opportunity may be less about giving businesses more marketing software and more about helping them delegate work they don't have the time or expertise to do themselves.
And if that means a small business owner gets to spend more time running their business instead of figuring out what to type into another prompt box, that's an outcome worth exploring.