AI Agents for Marketing Teams: Lessons from a 70-to-2 Business Rebuild

Key takeaways

• Mike Filsaime reports going from 70 people to 2 while quadrupling revenue
• In an agent-native model, people decide and agents execute
• Agents can cover support, email, social, video, funnels and development
• An agentic browsing score shows whether AI engines can read and recommend your site

In our latest video, we looked at one of the boldest examples of an agent-native business we have seen: Mike Filsaime's new project, which reportedly went from a team of 70 down to just two people.

This article breaks down what that model involves, what the agents actually do, and what everyday marketers can take from it, including a newer idea called agentic browsing that affects whether AI chatbots recommend you.

Disclosure: Prodify Digital offers a Scale.gg bonus bundle and may earn a commission if you buy through our links.

What happened when a 70-person business rebuilt around AI agents?

According to the figures shared by Mike Filsaime's team, over the last two years the business:

• Went from 70 people to 2
• Quadrupled revenue
• Cut running expenses 10 times
• Saw sales conversions rise 2.5 times

The point is not simply cutting staff. The business was rebuilt around AI agents, replacing the traditional bloated agency model with lean AI execution.

What is an agent-native business model?

In an agent-native framework, people and agents have different jobs:

• Two people make the high-level decisions
• AI agents execute the rest, around the clock

Because the agents are always on, work continues in the background after the team logs off for the night.

What tasks can AI agents handle for a marketing team?

In the business we covered, agents handle a surprisingly wide range of work.

For a smaller team, the sensible move is not to automate all of this at once. Use our guide to scoring AI use cases by value, risk and reversibility to pick the first one or two.

Do you need to be a developer to use AI agents?

No. As we explained in the video, you do not need to be a developer using tools like Claude Code or Codex.

You can run standard setups, or open an agent sidecar next to your apps and tell it what you want done in plain English.

That said, agents that act on your behalf still need sensible limits.

Our guide to using AI in a small business without losing control and our article on human approval checkpoints cover the essentials.

What is an agentic browsing score and why does it matter?

This is one of the most useful ideas from the video for anyone with a website.

Why SEO alone is no longer enough

Traditional SEO gets your pages read by Google. But many people now ask ChatGPT, Claude or Perplexity instead of using a search engine. Those AI engines also need to read and navigate your pages.

Optimising for them is called answer engine optimisation (AEO).

What a 3 out of 3 score means

An agentic browsing score of 3 out of 3 means AI agents can easily crawl, navigate and trust your site, which makes them more likely to recommend your business. If your site fails, AI chatbots may not mention you at all.

Mike Filsaime's team tested five of their own websites, including Click Campaigns and Scale.gg. They reported near-perfect page speed and 3 out of 3 agentic browsing on all of them.

Quick AEO checks for your own site

• Put a short, direct answer near the top of key pages
• Use clear, question-style headings
• Keep pages fast and mobile-friendly
• Make sure important content is plain text, not hidden in images

Building an agent-led marketing workflow? Our Scale.gg review shows how an agent-native campaign tool handles checkout, email and trials.

How can marketers get started with AI agents?

1. List the repetitive tasks your team does every week.
2. Pick one low-risk task, such as email triage or social post drafts.
3. Let an agent draft the work while a person approves it.
4. Measure time saved and errors for a few weeks.
5. Expand to the next task once you trust the first.
6. In parallel, check your key pages for AEO readiness.

Frequently asked questions

The bottom line

The 70-to-2 rebuild is an extreme example, but the lessons apply to any marketing team: let agents handle repeatable execution, keep people on the decisions, and make sure AI engines can read your site as well as Google can.

Related posts

• How to Use AI in a Small Business Without Losing Control
• How to Score AI Use Cases by Value, Risk and Reversibility
• Scale.gg Review: How Agent-Native Campaigns Handle Checkout, Email and Trials

Garry Knight

I'm Garry Knight, the person behind Prodify Digital. I write about email list building, email marketing, SEO, AI search and the tools that connect them. My aim is to make online marketing easier to understand, so creators and small business owners can make informed decisions about building an audience and keeping people engaged. Here you'll find straightforward guides and product reviews that explain what something does, where it fits and which limitations matter. The focus is on clear explanations and useful next steps—not hype, shortcuts or promises of easy earnings.

See Full Bio