Two Things You Can Learn Right Now: Using AI for Marketing, and Marketing to AI

If you're just starting to learn marketing, I want to say one thing upfront: at this particular moment, you can learn to "use AI for marketing," and you can also learn how to "market to AI."

A shift that is already happening

To understand the difference between these two things, you first need to notice a new capability AI has recently gained: AI has started talking to other AI.

This capability has a name: A2A, or Agent to Agent. In the old model, you opened ChatGPT, asked a question, and it answered. What's coming next is that your AI will go talk to someone else's AI on your behalf, comparing prices, scheduling meetings, filtering potential collaborators, and then returning a shortlist to you. Google launched an A2A protocol in 2025 specifically to pave the way for this, enabling AI agents built by different companies to find each other, delegate tasks, and collaborate.

I cover the mechanics in a separate piece. The one conclusion to carry forward here is: from now on, some of the decisions that affect you will be made by AI agents.

Your marketing audience now includes AI

At its core, marketing is about making the decision-maker aware of you, remember you, and choose you.

Historically, that decision-maker was a person. So marketing meant learning how to reach people, leave an impression, and get them to buy.

Now, that decision-maker has a new member: AI. When someone is too busy to search themselves, they ask their AI: "Find me someone who does this," or "Compare these three options," or "Recommend a consultant." At that point, the gatekeeper deciding whether to surface you at all is AI.

The target has changed, so there are now two distinct things to learn.

Using AI for Marketing: AI as a tool

The first approach is using AI for marketing. This treats AI as a faster instrument, using it to write copy, produce visuals, edit video, generate headlines, run ads, and analyze data. The audience is still human; AI simply makes the work faster and cheaper.

This is genuinely useful, and a lot of people are learning it right now. But to be honest, the barrier to entry is falling fast. Tools keep improving, interfaces keep simplifying, and before long, knowing how to use AI for marketing will be like knowing how to use presentation software. Being baseline-competent helps you keep up, but it won't create separation.

Marketing to AI: AI as the audience

The second approach flips the frame and treats AI as a new audience. Since AI will play a role in future decisions, you need AI to know you, understand you, and be willing to recommend you at the right moment. In practice, there are three things to get right.

1. Get found

When someone else's AI is searching for a person or a solution, can it encounter you in the data it reads? Do you exist somewhere it can actually reach?

2. Be understood

Once it finds you, can it describe you in one sentence, covering who you are, what you solve, and who you serve? Is your information organized and clearly structured, or is it readable to humans but opaque to machines?

3. Earn a recommendation

When someone asks a specific question, AI judges that you're the right answer and surfaces you. This requires you to articulate precisely the situations where you're the best fit, rather than saying something about everything.

Both are worth learning, but which is scarcer?

I'm not saying you can only learn one. Using AI for marketing makes your work faster; marketing to AI puts you in front of a new category of decision-maker. Both are useful, and you can pursue them at the same time.

But if you ask me which one is scarcer right now and worth acting on sooner, my answer is marketing to AI. The first approach has many practitioners already, and the tools keep lowering the bar for everyone. The second approach is something most people haven't even registered as a problem yet. Getting yourself organized in a form AI can read puts you one step ahead on an invisible recommendation list that most of your peers don't know exists.

How to start marketing to AI

You don't need to understand the technology deeply. Start with these steps.

1. Say clearly who you are. One sentence covering who you are, what problem you solve, and who you serve. This step matters for both humans and AI. If you can't explain it clearly, people won't remember it and AI won't understand it either.
2. Publish a version machines can read. Beyond the page designed for human eyes, add structure that AI can parse easily: clear heading hierarchy, structured data, and a plain-language description file for AI. The same content should feel natural to a human reader and be accessible to an AI reader.
3. Test it with AI yourself. Paste your URL into ChatGPT or Gemini and ask it to introduce you. If the introduction is accurate, you've been understood. If it's wrong or vague, you still have work to do. This is an inexpensive feedback loop you can run continuously.
4. Be specific about the situations where you want to be recommended. What question does someone need to ask their AI before you'd want to appear? Write those situations down. It's more useful than trying to stuff in a list of keywords.

What I do myself

This happens to be something I'm already doing. My AI business card and my knowledge site are both designed for humans and AI equally: a person sees a clean, readable layout; an AI reads a structured dataset it can parse, retrieve, and confidently recommend from.

Closing thought

Anyone can use the tools. Using AI for marketing will become more and more common, which is a good thing and sets the baseline for everyone.