0.1 Course Introduction: What You’ll Build and Who This Is For

[00:00:00] Welcome. This is the first lesson of this course, and I wanted to walk you through on what you’re going to walk away with, whether this course is for you, and the site that we’re going to audit, including what happened to it after the audit was done and everything was implemented on the site.

You have heard at least three, maybe five, maybe even 10 years of the advice of doing entity SEO, but you had no way of actually measuring it. There is plenty of advice on entity strategy, but what was really missing from a lot of the projects that I have personally worked on was the diagnosis, a way to measure where you are and whether a fix that you have implemented has actually worked on the website targeted for entity optimization in particular

Optimize for entities has been standard advice and almost none of it came with a way to check whether it worked. Maybe you have added some [00:01:00] schema, you mention a few related concepts, and then the client asks you, ” has this worked?” And all you have is rankings that move for maybe at least 15 other reasons.

What you are left to report is a list of things that you change, but you don’t have a before, you don’t have an after, and you have no way to prioritize what is actually left from your entity optimization strategy. what you will have by the end of this course is, first of all, a repeatable approach of auditing entities on your website with a score out of 15.

This is a baseline that you can re-run with a proven approach that works on

identifying the issues that might be hurting your entity SEO. You will also have a ranked backlog and a number that you can show and defend in front of your client. Everything else in the course is going to get you all of these four things

In this particular lesson, we’re going to cover [00:02:00] four parts. What you’re going to build, who the course is for, the demo website we’re going to use, and how this course is structured and how it’s going to run. Let’s start with what you’ll build. You will complete an entity audit on a real website and keep four things.

The first thing is a diagnosis, five dimensions scored zero to three each, for every page and for the site in total. you also have a ranked backlog, every fix scored on impact against effort and tagged with the pattern that it fixes, so you can hand it over to the team that is going to be working on it.

You will also have a roadmap, a 30-60-90 plan in an order where early work doesn’t need to be redone later

And the most important part is the toolkit. You have six browser tools built for this course and yours to keep: an extractor, a knowledge graph reconciler, a consistency checker, a gap matrix, the scorecard, and a [00:03:00] brief generator. All of those six tools are running in your browser, and your keys stay in the tab.

On top of these, there is the Claude Entity Analyst, which you will meet in module two. It picks up what Google’s Natural Language API can’t see with three checks for the entities a model makes up. The first audit teaches you the method

And if you know me, you know that I don’t like just talking about theory. So there are 17 step-by-step practicals, and several of them are automated with Claude. You’ll run the crawl through Screaming Frog’s MCP server and find competitors through data for SEO’s MCP server. In module five, you will learn how to write meta description at scale, run extraction checks, and build a link plan and a rules engine, and also generate JSON LD from an entity registry.

Every script, prompt, and plugin is public on the MLforSEO [00:04:00] directory, so you can read the code before you run it

Having a score in the audit is the entire point of having the audit. You can rerun it in 90 days and show movement. This is a number that you can put on a deck, and it is a number that you can defend. Entity work so far hasn’t had one, and I’m very happy to teach you how to audit entities for your site

Now let’s talk a little bit about who this course is for.

This course is perfect for you if you run SEO at an agency, if you are a freelancer, or if you work in-house. You have been told to do entity SEO and you want a measurable way to do it, and you’re also comfortable with working with data, spreadsheets, and you are enthusiastic about AI and AI powered automation.

It is perfect for anyone that not only has to rerun this audit on one website, but perhaps has to scale this as a service on an agency portfolio or [00:05:00] a client portfolio if you are a freelancer. So you don’t need any prior knowledge of entities or NLP. All of this is going to be explained in the course. You don’t need to be able to write the code yourself. Of course, we now have AI powered tools that can write the code for us, but it will help if you are a mid to senior level SEO and can troubleshoot at a comfortable degree.

You’re able to make decisions for the business and the brand based on the expertise that you have as an SEO

Whenever we have a practical that’s using a script that is written for you, so you don’t need to have a super big budget. And you also don’t need to have a massive budget in order to complete this audit. For instance, the demo website was just about 650 pages, and it really cost us nothing to run this audit with tools that I’m demoing in this course.

In order to learn this method, you also don’t need a large website and I’m [00:06:00] assuming you already know SEO and you know a thing or two about entities

Rest assured, even if you don’t, I’m going to walk you through all of it Sites up to 50,000 pages are the main path that we’re going to follow in this course from modules zero to five and then module seven.

Enterprise scale is optional. I’m assuming that not that many people are going to be working with large websites. But we still have module six that is going to cover entity pipelines in BigQuery and running the whole audit as an agent with Cloud Code. It is going to open up in the end of November, but it is still going to be included for everyone that has purchased this course

Things like structured data and internal linking get one lesson and a practical. Automating it at scale are separate courses that we’re working on to bring to you through the ML for SEO Academy. Building a knowledge graph and a brand strategy aren’t covered [00:07:00] in this course. This course is the diagnosis layer, while Beatrice Gamba course on building brand entities for AI search and knowledge graphs is the course that covers this part.

If you know that what you need is a brand entity program with digital PR and a knowledge graph that you publish and maintain, that’s Beatrice Gamba’s course, all 51 lessons of it. But this course is the one that tells you how big your gap is and which are the entities that matter based on the content that you already have on your site

Now let’s talk a little bit about the demo site that has been used throughout this entire course. This is a real website. We’re not auditing a fictional client. We’re auditing the MLforSEO.com and the academy.mlforseo.com websites. Those are my own, live and unedited. Entity audits look easy on a made-up example, but they can get hard on a real one [00:08:00] where the data is messy and the problems might overlap.

And my site turned out to be a great teaching example, even more so than I thought before I started planning this course There are 647 pages on it, 102,752 entity mentions, and it really cost me nothing in API fees to run the audit with the tools that I am demoing. It scored 5 out of 15, and you’ll see every point that it lost and why

Let me just bring you one of the findings that we encountered. The name ML for SEO appeared 148 times on this website, and Google’s language model typed it as a person 102 of the times, as other 36, and only 10 typed as organization. So essentially, Google’s models are reading my brand as a person seven times out of 10.

If you just add one word to this, ML for SEO Academy, [00:09:00] and then it becomes an organization all 92 mentions. Part of the reason it turns out was in the site’s own structured data

Every figure that you see demoed throughout this course comes from an audit snapshot that was done in the middle of August that comes together with this course. So your numbers, whenever you are practicing, are going to match mine and those in the slides exactly

So basically, the audit told me that Google knows the person behind the organization, but it doesn’t know the company. Lazarina Stoy, the entity behind the organization, resolved to a stable knowledge graph identifier, while MLforSEO had no entry on its own. So Google can tell you who I am, but it has no idea about the business that I run.

And because it’s my site, I didn’t just stop at the diagnosis, and this is what I teach in module five. I’m going to teach you how to fix these issues and measure every change.

One example is a [00:10:00] page about entity prioritization that the API read as being about the word entities. Salience of the main entity was 0.011, ranked third, and we have a mismatch between what we intend the page to be about and what the page is actually about. We did three rounds of editing that you can see in this lesson, and did the measurement after each round.

at the end, the subject ranks first at 0.231, and the verdict is a match. and on the same page, the brand schema node, which declared two types at once, person and organization, now only declared one. One of those round of edits only half worked, and you’re going to see the full process so you can plan your content edits better knowing exactly what moves after a certain content change using the tools that we have in the course.

Now let me walk you through how the course runs

There are eight modules, 48 [00:11:00] lessons, including the practicals. We are, of course, going to add more practicals if the learner community for this course requires it, or if there are any changes to the method as the industry progresses.

We first start with the setup, then foundations and the scorecard. We have three modules of diagnosis, extraction, the knowledge graph, and then competitors. Then we are focused on fixing and remediation. Module six, like I said, is all about enterprise websites and building this process at scale. And module seven turns all of those things into your 30, 60, and 90-day plan

We now have a pretty healthy library of courses on the ML for SEO Academy, so you might have some questions on where exactly this course sits. This course asks the question on what is broken. The Entity SEO and Knowledge Graphs course by Beatrice Gamba asks, what should we [00:12:00] build? The introduction to ML for SEO asks, how does this technique work?

You don’t need any of those courses to complete the other one. However, as you will find throughout this course, there are so many complementary learnings that you are going to get from Beatrice Gamba’s course if you are taking this one. So I highly encourage you to check out the bundles on the ML for SEO website, because we’ve tried to pair courses that naturally link together to help you make gains on your organic visibility through entity SEO or through AI automation

Every concept that is introduced in this course is defined before it is used. The audit runs on free browser tools and Google’s free tier, and practicals that need a paid tool are going to say so upfront. References are open access, and every script that you are going to encounter in this course is publicly [00:13:00] accessible. if this is the only course that you ever take from the MLforSEO Academy, it’s going to work completely with giving you all the knowledge that you need to execute this process.

The others, of course, do go deeper in their respective directions, but they don’t fill gaps in this course

Every lesson ends with a list of suggested resources, and all of them are free. You have definitions, you have short guides, you have different tools that we’re recommending for every lesson, and the course code and Slack community for questions

In the next lesson, we’re going to start with the setup. You have to set up two Google APIs on one key, plus one Anthropic API key with the billing enabled on both of these platforms

And then we can start

That being said, thank you so much for considering this course, and I’m extremely excited to walk you through everything that I have prepared for you in the AI Powered Entity [00:14:00] Audit course. I’ll see you in the next module