We've built the world's first ChatGPT simulator. You can see exactly how it picked its citations. Know exactly why your competitors are getting mentioned and you aren't.

For me, this release is less about “an API” and more about closing a long-standing gap in how SEO actually gets done. Most SEO work does not fail because of bad strategy. It fails in the handoff: insights live in one tool, writers work somewhere else, priorities sit in a doc,

While everyone was off at #TechConnect last week, I stayed heads-down doing what I love most: building. Ended up creating one of the coolest new features we’ve ever added to @mktbrew. Can’t wait to share it soon.

Don't think of LLMs as entities but as simulators. For example, when exploring a topic, don't ask: "What do you think about xyz"? There is no "you". Next time try: "What would be a good group of people to explore xyz? What would they say?" The LLM can channel/simulate many

SEO shouldn’t be guess-and-check. With @mktbrew, you can see how the model reacts before you push a single change. That’s where the industry is heading...we’re just already there.

Ok here's your chance to be my one thousandth follower! 🥳

“that’s just how search works.” It isn’t. With search-engine modeling, you stop guessing and start verifying. Every change you make is backed by a calibrated simulation of how the algorithms actually behave. No more publishing blind. No more waiting for “the next update.”

Our search engine model reveals thousands of these mistakes for our clients. It's so common to use H tags for styling. Don't! Use spans and CSS.

Behind every search query is a person. Behind every ranking is a reason. AI should bring us closer to that understanding, not hide it behind more complexity.

This week, I’ve been thinking a lot about what I’m grateful for. First and always, my family. The people who keep me grounded, inspired, and moving forward every single day. I’m grateful for the new team we’re building, a group of people who show up with heart, curiosity, and a

Thank you Claude! 🚀

8/ AI search isn’t a black box. It’s a system with patterns. Study the patterns, and everything gets clearer.

7/ The more the industry talks about grounding, embeddings, and model bias, the more predictable AI Search becomes. Exciting to see the field shifting from prompt tricks to actual structure.

6/ My favorite part of all this: everyone discovering simulation. You learn more from modeling the system’s behavior than from a thousand test prompts.

5/ Prompt tracking always felt like watching weather through a keyhole. Sampling noise hides the real story. Embedding relationships are where the actual bias lives.

4/ The move from CTR to “Selection Rate” is right on. AI models don’t read your whole page. They grab tiny snippets. If a paragraph can’t survive on its own, the model usually tosses it.

3/ Brand saliency = understanding what the model thinks your site is about. In practice, the embedding centroid tells you almost everything: • what topics you’re associated with • where you’re relevant • when you’ll actually get selected

2/ “Poking the black box” to see how the model reacts made me laugh. That has basically been my job description for years. We just called it calibration before it had a cooler name.

1/ Loving the deeper conversations popping up around AI Search lately. Mechanistic interpretability, brand saliency, grounding, selection behavior… Nice to see SEO finally talking about what actually drives AI answers.

I love this one. And now it’s just one of the four signals you can tap into inside Market Brew. Keyword Fueler takes what you’re already spending in PPC and turns it into real, compounding organic gains. Feels good when the system finally works with you instead of guessing.

One of my favorite lessons I’ve learnt from working with smart people: Action produces information. If you’re unsure of what to do, just do anything, even if it’s the wrong thing. This will give you information about what you should actually be doing. Sounds simple on the