Use Data Or Be Used By Data!

The September 21 issue of Seotistics is here for you!

I already discussed this topic in the past because it's one of the most crucial topics in Inbound Marketing.

Easy to talk about huge websites... but what about small ones?

There are key differences in terms of Analytics stack and strategy, for sure!

Please move this email to your Primary inbox or reply to it. This is to prevent Seotistics goes into spam by accident. Gmail users can read this tutorial to do it.

Read this in your browser

📣📣 Important Announcements 📣📣

Belgrade SEO was amazing, my slides and the ones for the other speakers can be found under Resources at the bottom.

I am still a little bit sick and tired from my trip, so you will see a surge in content by mid week or next week.

It took me quite some effort to product this issue today.

As for conferences/events, I will be present at:

SEO On Stage (Athens) - 24 October 2026 (contact me first for 70% off)
MeasureSummit October 2026
MeasureCamp Zurich - 21 November 2026
3 more I will announce soon but some already know 👀

As usual my latest course is here:

Become a Top Analyst Now!

Oh, and did I say I am a DataForSEO partner? It's my most used API and one of the absolute best for workflows.

You can start using it with my personal link.

Keep the above picture in mind because you will check it again once I start detailing some workflows.

What Changed In 6 Months

The stack I recommended before was GSC, GA4, GTM, Bing Webmasters, BigQuery, Dataform and Data Studio.

That list still holds but the layer sitting on top of it is what I missed.

The base stack looks like this:

Apart from the GSC logo and Looker Studio, the idea is the same.

But now you can simplify it further and unlock new ways to interact with your data.

n8n, an agent and a cheap LLM, all on a small VPS. This is a gamechanger because it's "cheap" and lets you do more with your data.

Once again, you can use something like this to actually act based on your content calendar.

If you sell products, you could also instruct AI to help you with your catalog 👀

Small websites play at a smaller scale, so most things you can do quickly for the time being.

And most importantly... you can monitor your competitors, their prices or even see what they changed.

Cheaper Automation For Everyone?

Look, at the start of 2026 I was actively discouraging AI for most tasks because the return clearly wasn't there.

Now, with models like DeepSeek, you would be dumb not to automate something!

These are my costs for last week only:

This is what powers my personal content system.

A small business is faster in working on new pages or campaigns compared to an enterprise website.

Your advantage is that you can target the weak areas of your competitors... an approach named flanking:

Can't compete 1 vs 1? Easy, then only target the weak spots.

For example, websites like Ahrefs, Semrush and any big B2B SaaS have amazing content.

But they don't cover most topics I do and they are slower compared to me.

This means that you can easily cover topics others don't if you are quick enough.

The Situation Today

Small websites (as always) keep facing tougher competition and saturation.

This is very much true in the Anglosphere but not so much in other countries.

I've handled websites all around the globe and the difference is stark.

BUT

This doesn't mean you can't do business as a small business or new player.

Dark Social (& Hidden Engagement)

This is the most important concept in Digital Marketing for me.

We are under the spell that everything can be tracked and privacy doesn't exist.

Well, we aren't able to correctly attribute or identify some traffic.

This often includes word-of-mouth and sharing content in private circles.

Yes, far from the reach of nasty LLMs...

If you've followed Seotistics enough, you know that I am against the obsession for accuracy.

Now with LLMs, the delusion got even farther because many think you can actually FULLY detect LLM traffic.

Reporting

OK... then how do we structure basic reporting?

It's fine to rely on GA4 and GSC from their UI but my personal advice is to explore some other options.

They aren't as complex as you think and will allow you for easier AI connection.

Make sure AI doesn't control your final output completely. For the rest, that's fine.

I don't even recommend dashboards if you just need to check traffic and info like revenue, really.

Whatever you can build with AI or is more comfortable for you, do it.

The problem of reporting at this stage is obsession.

Opening a tool every 5 minutes to see if your page gets traffic.

Then we have AI reporting, which is hard to gauge and that's why I recommend 2 workflows today:

Analyze fanout queries
Report on a per-page basis

There are some Chrome extensions to export your fanout queries more easily but you can also DIY.

For context, fanout queries are done by AI, not by people, aka it's not user prompts.

Still, it's additional data you can use for creating content or tweaking your nice product pages.

The example above shows some queries to get you an idea of where you stand.

You can also assess each page's performance, both here and in GSC's AI section (minus the grounding queries).

What I'd Refuse To Pay For Today

Small websites get sold the same enterprise stack as companies with a hundred analysts.

What I say here is a recap of what we just said, in case you missed some key details and need it straight out.

4 things you can skip:

3rd party SEO suites that sell you data. At most, use DataForSEO (affiliate code here) + GSC + BWT, most research nowadays is better done via scraping
Attribution modeling, because you don't have the volume to make it mean anything
Dashboards nobody opens, when a scheduled report answers the same question for free
Data governance theater, since you aren't a regulated bank and your data fits in one warehouse project

You just need the basics of consent for compliance and the rest shouldn't even beexpensive.

If you want to grow, you need action and quicker decisions. OK... it's not that simple but what you can 100% avoid is looking at numbers.