Use Data Or Be Used By Data!

The July 21 issue of Seotistics is here for you!

Google Analytics 4 is hated by many. You see a lot of posts on LinkedIn of marketers complaining about it...

True, the product is bad but with over (allegedly) 85% market share, what's the alternative?

So instead of complaining about it, I show you how you can exploit it.

People these days are GA4phobic, they are scared and angry at GA4... it doesn't have to be this way!

P.S. New article about small vs big websites out 👀

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 📣📣

The Seotistics Summer Sales are live with 30% off on all the products (check them here) with the code: SUMMER25. Offer valid until July 29.

P.S. Some of you got an even bigger discount 👀

I released an update for Analytics for SEO (course) and still have to update the other 2 products.

I am updating my Discord community to make it a reference for my readers.

Since social media have lower reach and emails can bounce, communities are the only way to ensure you get my message.

We do have a baby forum too now and I will try to organize my thoughts there.

Some Reasons To Use GA4

First of all, the market share so you will never struggle to find work.

It's the best choice out there if you don't know what to pick as everyone talks about it.

Second, the free BigQuery export.

Third, the biggest web ecosystem ever created and integration with GSC and Google Ads.

Need more reasons?

(And, FYI, my course covering GA4 extensively is indeed 30% off).

The Many Flavors

GA4 isn't only the interface you see, in fact you have many choices to get the data:

API (yes Looker Studio uses this one)
BigQuery Linking (the best)
Data Transfer (inside BigQuery but using the API data)
MCP (the usual LLM shiny tool)

GA4 can now be accessed via MCP, meaning you can chat with your data.

As you know, this approach doesn't change any fundamental issue with the tool per se.

It's just a shiny addition to have the same problems as before, with a different format.

The MCP just adds another layer of abstraction, it's not as accurate as writing SQL.

Code is reproducible and can be standardized.

We prefer the BigQuery Linking for all these reasons:

Where It Gets Tricky

GA4 is a complete mess to understand in terms of attribution and scoping.

It's definitely the hardest part of them all.

Many often pair metrics with the wrong dimensions and the results don't make any sense.

The GA4 BQ schema is also a real challenge as you need to deeply understand what's going on.

Countless agencies I've worked with did the following mistakes:

confused GA4 API fields with the BQ export
didn't even know it existed lol
attribution mistakes
metric calculation mistakes

So it is hard but many refuse to learn new things.

If you are a web specialist you must be aware of GA4 data and how to make the most out of it.

Otherwise clients will just choose a different provider...

or if you are an in-house, you will need to rely on someone else (which is always bad).

Where It's Better Than Universal Analytics

GA4 is much much better than UA for the BigQuery export and the event-centric structure.

I am not a super fan of this new structure but it's objectively better.

With some data modeling and understanding, you can easily extract info about your products and pages.

Event data is still super important because it allows you to see what happened at an exact point in time.

It's easier to think about your marketing efforts this way.

Company Perspective

Remember, if your company wants to switch out of GA4 it will need to find professionals who are capable of working with alternatives.

This means a smaller pool of potential candidates.

Apart from that, GA4 offers you free BigQuery exports, whereas the others make you pay.

It's too easy to be charmed by alternatives that overpromise to solve all of your problems.

They won't.

If you struggle to use a super documented tool like GA4, you will also struggle with any other tool.

Plus, are you sure this isn't another way to waste company budget rather than talking business and doing?

Professional Perspective

If you need to upskill in something, start from the mainstream.

Sure, you may have read of some other better alternative... but what's the market share?

I'd rather use a mediocre tool than being out of work.

Tools with more problems and/or complexity give you free work 👀

Just think about the amount of money made with GDPR and Consent Mode... quite a lot, uh?

I'd still go for GA4 because it's well documented by the community and you'll always find something to do.

Then, you can specialize into some other alternative that you deem more worthy and can recommend to clients.

The BigQuery Export

This option lets you export your raw data into BigQuery. Yep, Google takes care of it and you get the rawest data available.

This is massive actually.

The only downside is understanding its schema:

Learning this schema above will inevitably lead you to master GA4.

Once you figure out the inner workings of the tool, the rest will follow.

GA4 BigQuery is hard, don't let it fool you.

Is it harder than learning GA4? Absolutely not!

I'd do it over and over, learning how to play with data warehouses is key in Marketing Analytics.

Some Alternatives

For Product Analytics, Mixpanel and Amplitude are the absolute best.

Some companies think differently and choose to have both GA4 and an alternative.

This approach may work in some cases, e.g. you need specific data from a competitor but still need GA4 for the most.

I am not a big fan of it because it adds one more data source you need to account for.

Still, these are some competitors:

Matomo
Piwik Pro
Adobe Analytics
Triple Whale
Publytics

The list isn't even complete... it's better to understand your business model and choose accordingly if you really need to change.

How It Should Be Done

The ideal approach and what engineers do is to transform your data to a better format.

Instead of using the complex nested schema, we decompose the table into multiple ones.

Tools like GA4Dataform can massively help for simple projects where you just need the data out.

For more complex scenarios, it's time to do some data modeling.

I will spend more time on this topic in my next issues.

Does It Suffice?

No, GA4 alone doesn't get you a job.

I never talk about it as my only tool because you should never rely on one data source or romanticize it.

We are still doing Analytics, so we are by definition tool-agnostic.

Data-wise, this is what you also need:

GSC
GTM
Crawl data
Log files
CRM

Depending on the company, you may also integrate GA4 with your CRM, like Salesforce Marketing Data Cloud.

This unlocks new opportunities in terms of what you can monitor.

An example of stack for SEO but also for pretty much a lot of other tasks.

Skill-wise:

Communication and reporting
Domain Knowledge
SQL/Python to play with the data
Business/Marketing knowledge

My course "Think Like a Web Analyst" (30% off with code SUMMER25) explains you all of the business side of the job:

Become a Top Data Analyst Now!

The Metrics

I could start a war on this topic because the majority of the content out there is basic.

Many dashboards often report on standard metrics, such as:

Users
Pageviews
Sessions
Bounce Rate
Engagement Rate

I still read terms like "visits" or "traffic". These are not metrics, be more accurate in your language.

The outcome is that you get a dashboard conveying a lot of information but 0 action.

At a strategic level (i.e. you talk to the board), this is OK.

They just want information and you need to show off.

But if you have to create a dashboard to be useful, then you can't afford to show me Users and Sessions and call it a day.

There are other valuable options like:

Leads/Conversions
Custom events
Custom metrics
Custom dimensions

Every business is different so you'd better have custom criteria.

Avoid relying on the baseline metrics that don't make much sense for you.

(Still, there is a degree of standardization, we agree that revenue matters in Ecommerce, right?)

Commercial Value

Ok so... what's in it for you?

Well, learning GA4 opens many doors as I said and this has several implications:

you can serve more clients and offer more lucrative services
you don't need to ask another guy
better reporting in general

I still believe that it's impossible to report on SEO/PPC without GA4 (or any alternative).

GSC and Google Ads alone aren't enough.

And no, AI traffic has not made GSC obsolete as some misinformated post claimed on LinkedIn.

The 2 tools achieve (and measure) different things.

And even so, financial data (revenue) and lead data must be measured with external systems.

So no, GA4 is all but sufficient.

GA4 recently added lead reporting under "Acquisition" but not everyone has it.

My Seotistics property doesn't have it, another US property I own has it.