Welcome to the first edition of Web Anatomy Report.

Quick story before the first report.

I'm Gabriel Amzallag, and I'm a little obsessed with website growth (I have other passions too).

The short version of my background:

• 4 years at Qonto, a French unicorn, working on website conversion and SEO.

• Left in 2025 to go freelance. Since then I’ve advised 15+ companies on website growth, B2B names like Payfit, Pigment, Hiway, and consumer brands like Staycation and Hapik.

• Across every one of them, the same question kept coming back: which section do we use here? what structure converts? how do we rebuild this page so it actually works?

So I started curating the best sections and page structures I could find online.

This mattered before AI, AI just raised the ceiling.

Generic input in, generic output out. A model can build a better page, but only with the right references: My collection is that context.

So in January I turned it into a website with webanatomy.ai, 200 new sections and best practices curated every week, so anyone can give the right context when prompting.

And I decided to focus on what AI can't copy: curating with taste.

So whether you want inspiration on how website is evolving with AI, you’re sharpening your homepage and paid pages, or you’re just web-growth curious, you’re in the right place.

We’ll cover design, conversion, copywriting, and yes, SEO/GEO, because all of it is the website’s experience.

Let’s get into the first teardown.

1. The Ranking June 16–22, 2026

And today I will go in details into Texau homepage and their positioning around MCP.

2. Teardown of the Week TexAu

Most GTM tools have one buyer: the human who’ll use the dashboard.

TexAu built its homepage for two, the human operator and the AI agent buying enrichment through an MCP server.

And that’s what make it brilliant.

TexAu has one engine, but it sells it to two completely different people on the same page: the ones who want an interface, and the ones who want an API/MCP.

And I know a lot of companies are stuck in exactly this dilemma right now.

What makes it work:

1. Name the audience

The split is on the page, not implied.

• The H1 of the page literally reads “From a table or an API,”

• They named them GTM operators and GTM engineers, each with a clear CTA to its own product page (the platform, or the MCP page)

2. Let people see how it works

Animated demos show the product in motion for each audience, so nobody has to imagine it.

And that’s even more needed for the MCP audience that didn’t exist 18 months ago.

3. Show how fast setup is

Then a section shows how fast it installs — one command, copy-paste. That's a second style of CTA: not "book a demo," but "you're already running."

Looking for inspiration for your own sections? Go check my library: webanatomy.ai/best-landing-pages/sections.

What to steal: if you’re heads-down on your API/MCP right now, there might be a play here — name your two audiences, give each a mirrored “path,” and let visitors self-select on the first scroll.

3. One last thing

This was my first edition, so I’d love to get your feedback.

I’m also building Web Anatomy in public on X / Linkdin, and adding new sections every week on my website.

See you next week