Shai Aharony

Reboot Online SEO and Web Marketing

Managing Director

Flight Centre Travel Group

I was invited to help shape up a new in house online marketing team and to introduce the company to organic seo with the aim to decrease its reliance on paid search.

Online Marketing Consultant

Time Inc.

Senior Oracle Developer

Marketforce

Oracle Consultant

CACI International Inc

Oracle Developer

The proud MD of a fresh and innovative digital marketing company built on the most important and rarest of foundations; amazing people.  I freely admit that i possess an unhealthy enjoyment of disproving opinion based myths surrounding the SEO world with cold hard fact based data derived from meticulously planned and executed SEO experiments.  Thoroughly enjoy helping businesses and brands ensure they earn the respect of the Google algorithms and hence improve their visibility online.

Sika

- Producing business use cases for customer/marketing data
- Data Governance
- Data Modeling 
- Documenting processes

Data Analyst

- Web/SEO reporting and data analysis for the entire company
- Monitoring International SEO
- Creating a data system to store and report on Web/SEO data
- Data Governance 
- Keyword Research 
- Creation of SEO trainings

SEO Specialist & Web Analyst

Seotistics

- Turning web data into business value
- Creating robust processes and operations for your website
- Auditing and managing Content
- Using BigQuery, GA4 & GSC to craft actual use cases

Web Analyst Consultant

- SEO Content Strategy to drive Traffic for Displays Ads or Lead Generation
- Content Distribution and Repurposing to get the most out of your budget
- Helped an emerging entrepreneur to monetize their website via Lead Generation
- Data Analysis in Python/R to get insights on traffic and ads performance
- Increased clicks by ~15 times and CTR more than doubled (6 months) for 2 publishers

Freelance SEO Specialist & Analyst

Cesmed - Centro Studi del Mediterraneo 

I practiced with Joomla CMS and Excel and started to improve my knowledge in the sector

Search Engine Optimization Specialist

I am a Data/Web Analyst and turn data into actions, optimize processes and document best practices. 

My professional background in SEO, Marketing and Content helps a lot when using data and building processes. 

If you feel lost with your web data and want a solid action plan, contact me now!

Marco Giordano

My advice for those who want to learn GA4... is don't learn it! 

Or well, learn GA4 from its BigQuery schema like I did, because the UI is unusable for serious projects.

If you want to actually help your clients, you need to provide the best solution for the long term.

The UI (but also the API) is absolutely unable to scale: 

❌ Sampling, limits (quota limits, retention, cardinality, this and that)
❌ Actually having to use the tool lol
❌ Educating people to use the tool (good luck)
❌ Excessive discrepancies and too many details 

If you are an engineer, you don't even need to read this... you already know an interface isn't solid and scalable. 

The BigQuery export gives you the rawest data, aka you have full control over what you do with it.

(It also gives you mastery over the data, how metrics are calculated and what is what). 

But many marketers fixate with shortcuts and nice interfaces and lose in the long term.

So my advice is not to be discouraged by the bad talk online and just do it. 💪

There are many tools that can help you:

✅ A load of free resources online (including mine)
✅ LLMs like Claude doing the work for you. Even GA4SQL, nice website! 
✅ GA4Dataform making that mess of a schema usable with 3 clicks

Then, you can create whatever you want as always with Looker Studio or even Sheets. 

So, I ask you again, is it necessary to waste time on GA4 reports? (Spoiler: no.)

Just learn how to configure it properly and understand the key settings.  

📥 If you want to learn more about Web Analytics, subscribe to my newsletter in bio!

Data/Web Analyst | GA4, GSC, SEO, Content | BigQuery, Python, R, SQL

Competitive analysis is much more than opening a tool and exporting CSVs.  
  
Here I compare different domains to see how many keywords they share.  
  
I've taken my keywords in BigQuery (GSC) as a starting point to check the SERPs with DataForSEO.  
  
You can use this matrix to:  
  
↳ Actually understand who is a competitor of yours  
↳ Which competitors are similar to each other  
↳ Understand where to focus  
  
Once we have some more data (and less influencer hype), we could even assess similarity for AI Mode.  
  
❗️For the "top" keywords I don't rank since most of my traffic is branded/Direct/Social/Email but this helps me a lot to identify gaps for all the channels.  
  
P.S. Guess which domain is the outlier and explain why it's featured in the table 😂  
  
I will publish an article about SERP Clustering too this week (and much more), stay tuned.  
  
📥 If you want to learn more about Web Analytics, subscribe to my newsletter in bio!

Always check the clicks distribution of your website!  
  
🤔 How many pages are getting clicks and which ones?  
  
The normal SEO advice is to focus all of your attention on the best pages…  
  
or to find opportunity pages.  
  
This advice is good actually but it needs context.  
  
💡If you have a lot of templated pages, you can go the other way and apply changes at scale.  
  
The weight of Tech SEO may not be much for small websites but it stacks for bigger websites!  
  
Tiny changes aren’t tiny anymore if applied in bulk.  
  
The example below is a medium-sized website and it gets 12% of the total clicks from the top 1% pages by clicks (all non-branded).  
  
It’s quite diversified in terms of organic traffic if you think about it… the other 88% is given by the 99% “worst” pages.  
  
I know, talking about percentages may make your head spin but think about it…  
  
Is everything 80/20 as they say?  
  
Nope, you have to analyze case by case, although you will often deal with skewed data, that is true.  
  
P.S. The plot below takes the clicks by page (grouped data) and shows a rank on the x-axis, where 0 is the best rank (more clicks).  
  
🤔 Did you know of this method?  
  
📩 Want to learn more about Web Analytics? Subscribe to my newsletter in bio!

Single Page View, having all the data you need for a single page.  
  
Gather all of your useful data and merge it to create a single record for each page.  
  
This is essential for content auditing...  
  
and for advanced web analysis.  
  
You don’t want to look only at GSC data alone when deciding to do some pruning.  
  
GA4 data informs you about all the other channels to prevent you from making bad decisions.  
  
💡 As you may know, most of what you get in GSC by default isn’t super informative, you need to create your custom metrics.  
  
Once you figure them out, you can go wild and report more than clicks and impressions.  
  
Once again, content plans and CRMs add even more power to your analysis because you can add:  
  
👉 author  
👉 published date  
👉 last update  
👉 clusters/tags/categories/notes  
  
All of this is another way to label your pages easily… instead of measuring standard metrics.  
  
Any other data source is well appreciated if you can use it!  
  
The same goes for including dates or drilling down.  
  
In some cases, you may want to go deeper and see what happens by month.  
  
In the majority of cases, you'd only aggregate by page and set some conditions for pruning/optimizing:  
  
👉 below a certain performance threshold  
👉 by inlinks  
👉 by engagement  
👉 by $  
  
The possibilities are endless but what matters is the result!  
  
Fit this information into a process and you'll be much more efficient.  
  
P.S. You can also list all the queries a page ranks for in a single cell but that's a bad practice.  
  
📩 Want to learn more about Web Analytics? Subscribe to my newsletter in bio!

GA4 event data in BigQuery is hard to understand at first.  
  
Because it's actually a lot of tables and not just one, every day gets its own.  
  
Every row represents an event in the infamous events table.  
  
A parameter is additional info on an event.  
  
❗ The event_params column contains a lot of values inside it.  
  
It's an example of a nested and repeated schema.  
  
5 columns and just 1 row for all those parameters.  
  
It's like a small table within our bigger event table.  
  
The vast majority of website audits involve user interactions, so the events table is the MVP.  
  
🤔 The first_visit event has a parameter called page_location, namely where the event takes place.  
  
The URL is a string, so the value is under string_value and absent from the other 3 columns.  
  
ga_session number is an integer because it's a whole number. The other columns stay empty.  
  
You can do a lot with the events table:  
  
✅ Combine it with GSC to get additional page metrics  
  
✅ Sending your custom dimensions and properly classifying your pages  
  
✅ Explore how users navigate your pages  
  
✅ The good ol' analyses with Ecommerce data  
  
💡 The most effective exercise is grouping pages by performance or specific dimensions to give better advice.  
  
"Is this page among the top performers for revenue or is it only bringing traffic?"  
  
"What's the cluster it belongs to?"  
  
❗ Having custom dimensions to classify your pages is the best thing you can do so you'll get this data in BigQuery.  
  
This table has some perks:  
  
👉 it's as granular as it gets (good for accuracy)  
👉 you can define your attrib. models, metrics, you name it  
👉 you bypass GA4 quota limits, sampling and other limits  
  
GA4 data is shown in this pesky format for storage purposes but...  
  
good organizations will change the structure given by GA4 and create more tables out of it.  
  
It's honestly hard to write queries so engineers like to make it easier for us (hopefully).  
  
I don't want to get too technical, check my future posts!  
  
---  
  
So yeah, querying this event table requires you to:  
  
- stitch more tables together  
- unnest the event parameters  
- calculate pretty much a lot of metrics manually  
  
But there are "alternative" methods to make your life easier!  
  
📥 If you want to learn more about Web Analytics, subscribe to my newsletter in bio!

Seotistics Black Friday Sales are coming next week! 

Yes, this is the time of the year with the biggest discounts.

✅ Learn GA4, GSC and BigQuery
✅ Use battle-tested frameworks for Web and SEO data/analysis
✅ Get access to Python notebooks and SQL code 
✅ Dedicated 1:1 support and community access
✅ Integrate LLMs into your workflow to make your life easier 

Product updates coming soon too, stay tuned via my newsletter in bio 👀

You can prepare your wallets and check the link in the comments 🔗

It's the last day for the 40% discount on all my products. 

Use coupon code: BF2025. 

This is the best deal of 2025, take your chance now 👀

🔗 The link to buy is in the comments ⬇️

Analytics isn't (only) about data.  
  
It's also about:  
  
↳ decisions  
↳ business  
↳ speed  
  
The more you obsess over technical tools, the more you'll get punished.  
  
💡 LLMs are teaching us that and this is a great time to go beyond "tactics".  
  
Do whatever it takes to help decision makers make better choices.  
  
If you have a solid understanding of your industry, you'll have no problems to adapt.  
  
❗️Finally, you must be quick.  
  
Some decisions and projects can't wait forever.  
  
It's better to be wrong and pivot quickly rather than waiting for Godot (aka 100% accuracy).  
  
📥 If you want to learn more about Web Analytics, subscribe to my newsletter linked in my bio!

Classifying SEO Automation & Solutions With This Matrix.  
  
What should a business do?  
  
You can think in terms of complexity and business value.  
  
🙅‍♂️ A complex solution is hard to adopt and use in reality.  
  
Low business value is something you want to avoid even more.  
  
This means:  
  
👉 SEO Analysis because just analyzing GSC data adds a lot of value.  
  
You can find potential ideas, and what to optimize and it’s not even that expensive.  
  
Even if you have a lot of data, you can use BigQuery and do something.  
  
👉 SERP Clustering is one of the quickest ways to spot opportunities.  
  
I was able to find ideas and financial opportunities in just 10 minutes.  
  
Your keyword list should be good and you should perform some cleaning first but it’s super simple to implement.  
  
You don't even need to create your script, just pay for a tool!  
  
If you are big, you can consider working on your own tool.  
  
👉 Most of the stuff you see on social media falls under low complexity and low business value.  
  
A Python script alone doesn’t change your business or make you rich.  
  
The same goes for the AI Automation you see everywhere.  
  
A custom GPT alone isn’t that hard to reproduce and won’t make you stand out.  
  
👉 Overengineering is the other side of the spectrum.  
  
This is the worst you can buy!  
  
Super complex tools or solutions won’t benefit you if you can’t make them actionable.  
  
Always go for the simplest solution first.  
  
In reality, most of what is done in SEO can be solved easily.  
  
Actually, almost every industry suffers from the same issue...  
  
👉 Custom tools and good SEO Automation are what you need in the long term.  
  
You can’t base your entire strategy on quick wins because it’s not sustainable.  
  
💡 This is why you create tools to scale and automate processes.  
  
Do you really need to spend a week to build a report?  
  
Of course not, you’d be smart and use Google Cloud.  
  
In other cases you can simply have a Python script to help you save hours or even days.  
  
Automation can get complex quite easily at advanced levels but it’s where you also get more business value.  
  
If you consider all the time and cost savings, the numbers are astonishing.  
  
💡Go for the quick wins first and see if they can be replicated in different sections of a website.  
  
On an enterprise level, you can also think about regions and countries.  
  
Once you are mature enough, consider building custom tools or automation that benefits your business.  
  
Automating reporting via Google Cloud is one such example.  
  
All of this can be estimated in terms of ROI to assess their feasibility.  
  
P.S. The example here is inspired by the book written in the picture.  
  
📥 If you want to learn Web Analytics, subscribe to my newsletter in bio!

Cool, you now have your GA4 data in BigQuery but what about conversions (aka key events)?

😩 The GA4 export does NOT tell you which events are conversions.

So... how do we do?

Simple, we use the GA4 Admin API to pass this information 👀

✅ Then, you can flag which events are conversions or not, with a boolean column: true or false.

This is my favorite trick especially if you have multiple subdomains with their properties and with specific conversions.

In all the other cases, you can simply flag them manually once you report on them, no need for this.

That is, if you actually have a few conversions, as it should be! 

P.S. Ofc you'd need to do this in Dataform/DBT and make it a little bit more "elegant".

📥 If you want to learn Web Analytics, subscribe to my newsletter in bio!

LLMs like Claude help TREMENDOUSLY with coding, especially SQL and Python.

Don't limit yourself to "I am not a technical person", it's no longer an excuse.

It's not like you need to memorize code or lines.. forget about it.

✅ Use your brain for actually giving context into data.

As long as you don't do fatal mistakes, it's 100% fine to rely on LLMs. 

You are paid to provide value and give a meaning to data, LLMs don't do that.

💀 Or well, they do it but it's mid, LLMs are simply your tools. 

Let's be real, most of the requests we do are always the same.

Document the code and automate what's possible! 

💪 Prompting per is quite simple for 99% of the use cases we have...

but I still see common mistakes like being generic.

A few days ago I also showed you to use a simple method to have more accurate answers: using sample .md files. 

Yeah, it looks simple and easy but it works.

Don't be lazy and give it a try 👀

🧪 I tested a LOT and I proudly say Claude + Cursor are my daily bread. 

I never learned Javascript but it's so easy to create some quick web app now.

And don't even think I write 80 lines of SQL manually 😂 I use templates though! 

📥 If you want to learn Web Analytics, subscribe to my newsletter in bio!

taplio

The Battle

Shai Aharony

Marco Giordano

Followers

Posts

Engagements

Likes

Performance ScoreMarco Giordano's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Shai Aharony

Marco Giordano

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Marco Giordano's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreMarco Giordano's LinkedIn Performance