The Seotistics newsletter is written by Marco Giordano, a Data/Web Analyst with the goal of combining business and web data. Tired of the usual boring Analytics content without any business impact? Seotistics teaches you how to use Analytics, web data and even content in your workflow while helping you with Strategy.
Share
π― Business Approach For Web Analytics
Published about 2 months agoΒ β’Β 7 min read
β
Use Data Or Be Used By Data!
β
The February 3 issue of Seotistics is here for you!
Boring works, in a world of hype, extreme focus and dedication are what get you results.
This doesn't exclude innovation or avoiding automation because you need a healthy dose of exploration.
This is referred to as Exploration vs Exploitation and is one of the key points in any organization.
And yes, this issue will cover a lot of BUSINESS.
If you love this topic you are lucky... I will fully cover it in my next course!
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.
"Think like a Data Analyst" will be my next course, as already said before.
N.B. The final name may change.
The program will contain:
How to think about problems
Managing a Web Analytics project from a business standpoint
Framework and processes to tackle business problems with web data
If you are an engineer, you will learn about business and how to use data for decisions.
If you are a marketer, you will enhance your knowledge further.
If you are an analyst, you will learn even more!
Those who voted for it will be contacted since I want to ask for their opinion again.
Why You Need Boring Web Analytics
Most of the effective things we do with websites are boring and repetitive.
Ecommerce is a prime example as most of what you can do is technical or involves organizing products.
Content websites are the most exciting of the bunch but require a high degree of standardization if you want to scale them.
Anyway, good work requires standardization and can't be defined as very exciting if you have been doing it seriously.
Procedures and standards are what get you through
It works, you can't start questioning everything
Analytics is invaluable at this point and it doesn't limit itself to Google tools, and please check my previous issue too.
β
π Good VS πΏ Bad Exploration
Exploring automation, coding, Analytics and cool ideas is fine if properly done.
I myself am a big fan of this stuff as you know.
The problem is when the following happens:
You obsess over AI because you fall for scare tactics.
You think that Python/Coding = Insights. Unfortunately, you need much more...
You explore too much and forget about your goals...
This lack of focus is harmful to your business and we have seen a recent example with Google Analytics 4.
While I agree that GA4's launch was problematic, not every business needs to be super proficient in it.
Exploration is the starting point for innovation. It requires you to think out of the box and explore new alternatives.
Exploration doesn't need to lead to perfect results, just don't abuse it!
My recommendation is to test some innovative ideas and see how they fit into what you already do.
π Exploitation
Exploiting your knowledge is how you get through and the ultimate proof of competence.
Most of your efforts should go on using your existing resources.
If you aren't able to capitalize on the present, you will have a hard time in the future.
An example of knowing your craft is fending off competitors' attacks and eating their traffic.
β
If your processes don't make you competitive, then what's the point?
Most of my work involves using data more efficiently than what you see elsewhere... but the ideas are the same.
Exploitation is productivity and efficiency. Your comfort zone, your ideal playing field.
Making Order Out Of Chaos
Before you touch anything, figure out what the company already has:
dashboards
documentation
data pipelines
software
metrics/KPIs
The #1 mistake agencies make is rushing to execution without asking first!
Countless times I've seen external partners building awesome dashboards that just don't deliver.
The same goes for software, pipelines and any other output you can have.
I want to introduce you to this nice framework I just thought of:
Everything will be explained here. Metric trees and other cool stuff in my past issue.
It's common to replicate dashboards that already exist across the organization...
or even worse, have 10K models in DBT and you don't know what does what.
The lesson is to stop at what you need and not ingest too much data or have impossible requirements:
β
Having more data can either create more noise (and drop the value) or even make no difference.
A Simple Story
One of the biggest projects I've ever worked on involves building a data platform for a large enterprise.
As an Analyst, I cover the part involving business, data modeling and naturally, processes and documentation.
If you approach a project like this in terms of technology, you will fail.
Stakeholders are hungry for results and want stuff quickly.
Yes, because you start losing focus on what matters and talk about details.
β
Leave the technicalities to engineers, you are meant to do business!
Anyway, the first thing I did was to understand which use cases we could produce.
"How can we use Web Data to build something useful?"
The answer was simple, we needed dashboards this time because there was no reporting system available.
Imagine a company where you can't even report on data for all of your websites...
this is more common than you think!
This is why I recommend building something like this:
β
You can fill the solution section either later or immediately if you have enough expertise to find the solution π.
Modeling & Logic Model
Companies really struggle with having the data in the right format and understanding which metrics they need.
Data modeling is a nice skill to have that isn't just technical because it also allows you to describe the business.
For example, you can consider the following entities:
URL
Channels
Devices
Countries
Sessions
Events
Users
The idea behind data modeling is to represent how to store data and define the logic of how to access it.
I want to show one of the most basic examples, star schema with GA4 data (events table):
β
The central table is the fact table, the concept, the measure aka quantitative.
The other tables are dimension tables and complete the fact, aka qualitative.
Every dimension table has a 1-to-many cardinality. β In English, it means that 1 observation in those tables can be tied to more rows in the fact table. β E.g. 1 user can have multiple events!
But this isn't the only approach... recently event data modeling is becoming more popular.
There isn't one universal approach to how you represent data
...but it's important you know how your datasets work.
Based on the questions you often ask, you can change the schema you will use.
Don't forget to use this splendid framework by the Kellogg Foundation, i.e. the logic model:
β
It's a way to understand how your organization does its work and is able to link activities/processes to outcomes.
Things Can Still Go Wrong
Recently, I had to test different tables and verify that pipelines were working as intended.
Spoiler: they were not.
I noticed that GA4 numbers weren't matching and had to discuss with engineers...
In reality, most projects are quite complex and you don't always have the political power/resources to prevent problems.
For this reason, I recommend you only ingest the data you need and not everything.
Luckily, ingestion is not our responsibility and I leave this fun topic to engineers π€£
But as a marketer/analyst, you must move on and be sure that what you propose has a business purpose.
If you want to learn this for SEO data and GA4, I've prepared the ultimate course for you:
[Analytics For SEO Course - v4]
If you want to be guided and start from scratch, this is the course for you!
You will:
β Use GSC and GA4 Data to their fullest potential
β Learn Python/SQL for your needs
β Get a complete blueprint for auditing websites
β Learn how to 10x your productivity
β 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!
The Seotistics newsletter is written by Marco Giordano, a Data/Web Analyst with the goal of combining business and web data. Tired of the usual boring Analytics content without any business impact? Seotistics teaches you how to use Analytics, web data and even content in your workflow while helping you with Strategy.
Read more from Seotistics - Web Analytics + Business + Strategy
Use Data Or Be Used By Data! The March 31 issue of Seotistics is here for you! It's undeniable that the Western data job market has changed and is still changing. OK cool but what's your point? Simple, it's necessary to go beyond labels (e.g. SEO, Product, Growth, Analyst) if you work with web data. I've writte a new article covering the other part of today's NL, it's about Descriptive Web Analytics and why it matters. Please move this email to your Primary inbox or reply to it. This is to...
Use Data Or Be Used By Data! The March 18 issue and 100th of Seotistics is here for you! Finally, we are at the #100 issue! Thanks to all for the support and for reading my content with passion. In this issue I will go through our story and what you can expect to see in the future! 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 π£π£ Important...
Use Data Or Be Used By Data! The March 17 issue of Seotistics is here for you! If you have wondered aimlessly searching for cool automation through the depths of social media... don't worry, I will show you some ideas to get started and what you need to know in 2025. Consider this as a nice introduction to the topic. P.S. I am also working on a lot of new content for you folks, stay tuned π You will get some freebies and more stuff on what I'll discuss today. Please move this email to your...
