Use Data Or Be Used By Data!

The October 19 issue of Seotistics is here for you!

I am just back from Rome and AI didn't sleep, we are full of news and updates.

As you know if you are working with automation, you need to know this stuff!

As I try to recover from multiple conferences in a row, here is the latest summary of what's going on.

Most of these topics will be expanded in the next issues as they are too new 💀

​Read this in your browser​

📣📣 Important Announcements 📣📣

As for conferences/events, the next ones are:

• ​SEO On Stage (Athens) - 24 October 2026 (SEO-On-Stage-Marco70 for 70% off)
• MeasureCamp Zurich - 21 November 2026
• ​Found Conference Tokyo - 18-19 Feb 2027 (Super Early Bird active now)

I've also decided to add a weekly news recap in every issue.

The first few times it will be simplistic but I am curious to experiment more here.

There will be new cool changes to this newsletter and plenty of content updates 👀

News Recap

From October 12 to 19:

• ​BigQuery made AI.AGG generally available (October 13)
• ​Gemini 4 Argon went generally available (October 13)
• ​Claude shipped OpenTelemetry export in beta (October 14)
• ​OpenAI confirmed bug on missing UTM parameters (October 15)
• ​DeepSeek released V4.1-Flash-1017 (October 17)
• ​DeepSeek opened its Search and Fetch APIs (October 17)

I will comment each of them, don't worry.

It's a lot of technical stuff to digest, the actual English explaination is easy.

As you may notice most news are about LLMs and AI since it's a big chunk of our work today.

The DeepSeek Earthquake

DeepSeek is a major LLM player and one of the main reasons why US models should be scared.

2 main news:

• Search and Fetch APIs
• DeepSeek V4.1 Flash 1017

DeepSeek built its own index and search engine, so with only $0.75 you can get 1000 calls.

But... there is a catch here.

Search is one thing, Fetch is another one. Using both would cost you $1.50!

Search gets you the web results, aka the links.

Fetch gives you the stripped content of said links.

This is more convenient than many other services like exca.ai!

Expect these APIs to change how we automate some web retrieval processes.

Mind you, it doesn't replace services like DataForSEO, used for Google or other search engines.

And the biggest surprise is that:

• billing is per successful request -> empty result or failures cost $0
• you can get as many results as you want at the same cost (1 or 50 is the same price!)

This is dirt cheap.

Now consider that until October 31, this API is free within DeepSeek Harness, Claude Code and Codex.

You have 2 weeks to toy with this amazing service and see for yourself what's the output.

DeepSeek-V4.1-Flash-1017

The second biggest news is that they just released a model that isn't too far from Opus 5.5 at a fraction of the price.

I really don't understand why people use OpenAI/Anthropic models for simple tasks.

Unless you are doing nuclear physics, there is no need for that much power.

When building automation like I do, the key is to route different models and use the cheapest when needed.

Google Waking Up?

Google recently released Gemini 4 Argon, their latest LLM.

2 per million input and $10 per million output, doubling after Nov 30, and it's what AI Mode answers with for Pro and Ultra.

Despite that, it's not as good as other models... so it doesn't bother us much.

The only important takeaway here is about Google AI Search changing model.

Anthropic Riding The Wave

Claude now supports a telemetry export, in plain English you can see what an agent does and "thinks".

Debugging is now easier and you have the option to understand what's actually going on.

This is one of the most welcome additions, actually.

I am not currently using Anthropic models as I am loyal to DeepSeek.

Oh and I almost forgot, Opus 5.5 got Fast Mode, which simply means you can speed up the output 👀

BigQuery is an AI Powerhouse

Apart from storing all the Google data (GA4, GSC, Google Ads), this warehouse supports AI functions.

The use case for these functions is that you don't need any extra script or intermediate steps.

Another reason why BigQuery is my favorite warehouse.

The process to set them up is a little bit annoying though because:

• you need IAM (aka permissions)
• enable Vertex AI (which costs money)

Not a blocker for analysts but can be quite cumbersome for marketers who want to play with them.

Bright Future Ahead

I have quite some work to do on my content system and consider newer updates.

AI or not, you still need to spend time on your projects to make them florish.

Last week I showcased Jev (which is peak) and now we even have new LLMs to play with.

After I am done with Athens, I will keep releasing all the content, course updates and whatever I promised 👀

Including my job market analysis for Web Analytics, which is currently in progress :3

So, stay tuned and until next week!

👥 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.

I will start posting more there as we have a forum channel now.

This is the best way to stay updated in real time on Seotistics:

🔎 Analytics For SEO Ebook - Course / Ebook

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!

Also in ebook:

Think Like A Web Analyst

This course teaches you to:

✅ frame Analytics problems

✅ understand which metrics matter

✅ managing Web Analytics projects successfully

📚 Recommended Reads - Peak Content 🗻

Read these peaks (again):

• ​The Downstream Impact of ​​AI Visibility​
• ​Understanding Observability in the world of AI search data​
• ​Ultimate Content System: Scaling Seotistics​
• ​Orchestrating Systems & Automations​
• ​The Evolution Of The Web Analyst​
• ​Setting Up n8n On A VPS: A Practical Guide (Hetzner + Docker)​
• ​Hermes Agent On A VPS: The Complete Setup ​​Hermes Agent On A VPS: The Complete Setup​

As usual, my most recent LinkedIn content is here.

❗️ 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 167, Zurich, Switzerland
​Unsubscribe · Preferences​