Use Data Or Be Used By Data!

The February 17 issue of Seotistics is here for you!

There is too much misinformation on AI and this issue is dedicated on dispelling many myths you read online.

Agencies and companies need to know before they allocate budget incorrectly 👀

AI is an important aid for Web Analytics but people overdo it and this makes us worse.

​Read this in your browser​

📣📣 Important Announcements 📣📣

My new course preorder will come out soon with a 30% discount... hopefully next week!

I've created a LinkedIn page for Seotistics, go follow it!

brightonSEO released the video of my October presentation, go check it out!

I will be speaking in brightonSEO (again) on April 11 (Friday) about Content Auditing for small and big websites.

AI Overviews, Google Hype, "Research"

One of the topics last year was Google VS AI, especially ChatGPT.

In terms of measurement, nothing changed.

Unfortunately, I recently read that some people are changing their way of measuring organic performance...

and it's pure nonsense.

Remember, changing the way you measure things implies you have new data or a solid plan in mind.

Your most common traffic metrics are always the same and you have too many options:

A good exercise to do on your website is to measure your AI traffic from GA4.

As some of you know, this information is stored in the Page Referrer, so you can simply list down all the values.

These are some of them, like listed in this article by Two Octobers:

BigQuery users, page_referrer is an event parameter. If you wonder what I am talking about, you can check this article about GSC and GA4 data in BigQuery.

Publishers and content producers should be careful, as this study from Authoritas showed.​

It's clear that informational content isn't exactly the most defendable asset...

and this is why I encourage you to embrace Content Distribution & Repurposing ❤️.

The State Today

As of now, there is no real action you can do to rank on ChatGPT or even AI Overviews other than the usual SEO advice:

• Backlinks and brand mentions
• Great long-form content

And nope, Schema doesn't help as LLMs are not trained in that way!

Sure, we could speculate about it... but I'd rather stick to what is most likely.

Speaking of other marketing channels, there are a lot of nice ways you can employ AI:

• brainstorm ideas for content distribution
• generate or edit pictures (you must be good at it, though)

If you want a good summary of why AI doesn't actually think, I recommend this article by Juliana Jackson and this evergreen classic by Jess Peck.

Remember, Analytics is about simple stuff, not about cool complex stuff...

Seotistics teaches you to care more about the outcome/business rather than technicalities.

The Rebirth Of Make and n8n

The extensive use of AI tools made automation solutions like Make and n8n even more popular.

For example, you can create a workflow to understand when to update content...

It's nothing new but I recommend you look into it, AI or not.

My friend Katarina Dahlin has a nice video explaining to you how to automate product descriptions with Make and ChatGPT.

What Many Organizations Are Sleeping On

The usual stuff I talk about is what makes a difference:

• ask the right questions
• frame problems
• tie data to the business

Most of the problems today derive from imbalances in data work compared to the business.

Capturing all the data you can because "you never know" is one of these mistakes.

I see them in every organization so it's important to get these sorted out.

The Data Warehouse Lifestyle

Whether you do AI or Analytics (or both), you've probably heard me talking about having your data in one place.

This isn't some minor detail, it's key to being efficient and making your life easier.

If you work with Google APIs, you will soon realize that it's not worth it at all!

They suffer from several issues like:

• showing incomplete data
• quota limits
• confusing sampling

BigQuery solves this issue quite easily for both GSC and GA4... but further attention is required with Google Ads.

Speaking of Matomo, Piwik Pro, Mixpanel and Amplitude, you can also export your raw data to your favorite data warehouse!

Mind you, unlike GA4, you have to pay.

But the tool isn't that important, the concept is that you can do it from pretty much every good GA4 competitor.

Writing The Requirements

As said many times before, engineers may do the ingestion process and work on the technical stuff but...

you need to give them a direction on which data is needed and why.

If you fail this part, you are doomed because you will measure and produce useless stuff.

Being able to calculate the metrics and figuring out actual use cases...

my existing course shows you how to do this and beyond with SEO Data:

[Analytics For SEO Course - v4]

If you want to be guided and start from scratch, this is the course for you!

You will:

✅ Use GSC and GA4 Data to their fullest potential

✅ Learn Python/SQL for your needs

✅ Get a complete blueprint for auditing websites

✅ Learn how to 10x your productivity

✅ Learn BigQuery to work on large websites

I teach you what's needed to go from 0 to a professional Data Analyst.

Even if you leave SEO, the foundations are the same for other jobs!

AI & Data Analysis

I've seen many people using AI for Data Analysis but the answer is always the same...

AI should NOT be used for analyzing private data.

The goal of analysis is to simplify the initial complexity, i.e. you have data but no clue.

The better alternative is to ask for the code and adjust it from time to time.

But if you can formulate the right questions and know what to do, the code only acts as a means to achieve your goals.

I use Claude Sonnet all the time and it's quite good to complete my coding skills.

Active Challenges

AI has caught my attention for some specific tasks that burden SEO projects:

• Classifying pages
• Summarizing websites
• Internal Linking
• Extracting entities from text

I am actively working on solutions and I've made some good progress on some fronts.

We finally have access to new technologies so solving or improving page classification would be a massive win for me.

Summarizing websites is already feasible and I like the results so far!

GLiNER for example is one of the best solutions so far to extract entities. This wasn't my idea as my friend Andrea d'Agostino wrote an article on it.

GLiNER is not based on LLMs as the original paper states.

For internal linking, the actual challenge is being accurate. Tools like LinkWhisper do a great job but have several inaccuracies.

My (Very) Personal Use Cases

AI for personal use is overpowered, for example:

• Write code that took me hours, if not days, in a few minutes
• Prototype ideas
• Create diagrams faster

Small things stack and make you more efficient.

On an enterprise level, AI use cases are much different:

• summarization of documents
• chatbots
• structuring messy data
• clustering

Some of these topics will be covered in the next issues... big things to come!

👥 Join Our Community

Our Discord community offers a small place where we can talk business and web data.

If you hate all the noise of social media, then this place is for you.

🔎 Analytics For SEO Ebook (v8)

If you want to learn about Analytics for SEO and prefer self-study, this is for you!

It will teach you or your employees to:

👉 Prepare audits that make sense and are actionable 🔥

👉 Avoid common pitfalls that cost you money 💸

👉 Create analyses that add value and money💯

👉 Move the needle faster with efficient SEO systems ⏳

This comes with periodical updates to keep the content fresh.

📚 Recommended Reads - Peak Content 🗻

Some nice reads in addition to all the links above:

• ​Analytical Skills for AI and Data Science : Building Skills for an AI-Driven enterprise​
• ​Data Literacy In Practice​
• ​Becoming a Data Head​
• ​✅ Making A Web Data Project Successful In 2025​
• ​Solomon Kahn's newsletter - great newsletter about data leadership

My LinkedIn Content:

• ​GSC VS GA4: All the differences​
• ​Big websites have a lot of data​
• ​5 underrated analyses to do with GSC and GA4​
• ​Common and wrong data stereotypes​
• ​From Data to Action: DIKW Pyramid​
• ​Explorations in GA4 have their own share of issues​
• ​Minimal Analytics stack for small websites​
• ​Web Analytics isn't only tagging or technical stuff​
• ​GA4 + GSC data
• ​Tagging pages - Different methods

❗️ Feedback and Recommendations

If you have ideas/recommendations for the next issues of Seotistics, you can simply reply to this email.

Marco Giordano
​Data/Web Analyst

Follow me on 🔽🔽🔽:

​​​

Bernerstrasse Süd 169, Zurich, Switzerland
​Unsubscribe · Preferences​