How To Go From Web Analytics Specialist to Data Engineer?
The journey I took to go from debugging website events with Google Tag Manager to automating python scripts for data pipelines.
I started tracking websites some time ago. My usual tool stack includes Google Tag Manager, Data Layers, Google Analytics, and Looker Studio (formerly Data Studio).
Every now and then, I would use Apps Script on Google Sheets to automate certain tasks, and SQL was primarily reserved for implementing custom logic in Data Studio reports.
Nowadays, I work as a Data Engineer focusing on workflow automation, ETL, orchestration, data capture, and even AI-driven workflows for use cases such as RAG.
If you are labeled as a Web Analytics Specialist (or sometimes, albeit incorrectly, as an Analytics Engineer), I am here to tell you that there is not much required to switch your career.
There are many things you might already do that can be easily transferred to the Data Engineering mindset to make your career switch smoother and more enjoyable.
Let’s go through them!
📊 From Dirty Looker Studio SQL to BigQuery
If you have ever tried to do something advanced in Looker Studio, you might have encountered limitations at some point.
Blending data can harm performance if not executed properly, so you need to develop more complex approaches.
Using SQL is one such approach, as it allows you to define calculated fields easily with your own logic. For example:
CASE
WHEN session_count = 1 THEN "New"
WHEN session_count BETWEEN 2 AND 5 THEN "Returning"
WHEN session_count > 5 THEN "Loyal"
ELSE "Unknown"
ENDThe good news? You are ready to jump into BigQuery because the SQL syntax and nuances are nearly identical.
BigQuery, for any Google Stack data user, is the gateway to powerful data processing.
You no longer have to set up awkward Google Sheets connections to Google Analytics or Looker Studio if you have BigQuery.
You also do not need to spend hours debugging events with Google Tag Manager’s live view or other debuggers, hoping that events are being correctly captured, when you can simply confirm this by querying BigQuery GA4 tables.
If you cannot use BigQuery and are still reliant on Google Sheets-based reports, you can learn SQL with DuckDB using its gsheets extension. This will enable you to learn SQL and query data without needing to become an Excel expert.
📒 From Tracking Guide to Data Modeling
At some point, you may have designed a Data Layer for developers to implement when the page loads or when a specific form is submitted.
For that purpose, you would have analyzed event definitions, reviewed the website or app behavior, traced all event chains, and determined what occurs and where.
With all that context and the business requirements in mind, you would compile a tracking guide for the IT team. For example, you might send something like this:
dataLayer.push({
event: "productDetailView",
ecommerce: {
items: [{
item_id: "SKU_12345",
item_name: "Red Sneakers",
price: 89.99
}]
}
});Surprise—you were already performing Data Modeling.
Data modeling is the process of organizing raw data into clean, structured tables that answer business questions.
If you have ever built a dashboard in GA4 or set up a custom report in Looker Studio, you have already considered:
What dimensions and metrics matter (e.g., product name, price, revenue)
What level of detail you want (e.g., per product? per user? per session?)
How data should be grouped, filtered, or joined
That is data modeling—just accomplished in SQL instead of through dropdown menus.
Tracking is about what you send. Modeling is about how you prepare what was sent so it is easy to use.
Until now, whenever you were preparing a tracking guide, you were:
Defining schema, entities, and attributes (e.g., item_id, item_name, price)
Defining the grain of the table (e.g., User > Event > Item)
Determining which nested objects needed flattening so items could be listed properly (e.g., ecommerce.items into item_id, item_name, and price)
The previous example could be translated to:
SELECT
event_name,
item.item_id,
item.item_name,
item.price
FROM `project.dataset.raw_events`,
UNNEST(ecommerce.items) AS item
WHERE event_name = 'productDetailView'
🦸🏻♂️ From Data Layer to APIs
Let’s look how similar the concepts behind parsing data layers and API calls are.
A typical Javascript Data Layer coming through:
dataLayer.push({
event: "view_item",
ecommerce: {
currency: "USD",
value: 129.99,
items: [{
item_id: "12345",
item_name: "Wireless Headphones",
item_category: "Electronics",
price: 129.99,
quantity: 1
}]
},
user_id: "u_6789"
});A python API call with its response:
import requests
response = requests.get("https://api.store.com/products/12345")
data = response.json()
print(data)
# Response below
{
"event": "view_item",
"ecommerce": {
"currency": "USD",
"value": 129.99,
"items": [{
"item_id": "12345",
"item_name": "Wireless Headphones",
"item_category": "Electronics",
"price": 129.99,
"quantity": 1
}]
},
"user_id": "u_6789"
}It goes without saying that this is an example for demonstration purposes, but if you pay attention, there is no real difference.
Start exploring Python as a tool if you have not already; you can elevate your report generation and workflows to the next level.
Simply adding Python to your skill set will make you a Web Specialist with enhanced capabilities; you do not even have to switch careers.
You can begin creating your own scripts on Google Collab and interact with the Google Analytics API or any other source you are currently tracking.
🔗 From AppsScripts to Workflow Orchestration
This section extends the previous discussion to another concept.
At some point, you may have needed additional functionality and used Google Apps Script to develop more complex scripts on Google Sheets, executed daily, hourly, or weekly.
If I am correct, you were resourceful enough to extract data from an external source that tools like Supermetrics or Porter do not support, necessitating a workaround.
The script was extracting data daily, processing it, and saving it into another Google Sheets tab, which likely ended up in Looker Studio.
These are two of the most fundamental principles of Data Engineering: Workflow Orchestration and ETL (Extract, Load, Transform).
This is not a call for you to stop using these approaches, but to explore and expand how things can be done.
Entering the Google Cloud ecosystem allows you to schedule SQL queries and enable Cloud Functions using Python, which can then be consumed by BigQuery.
Google Sheets has saved us many times, but once you have millions of rows or need to work with sensitive data, it should not be so accessible.
⚒️ Use NoCode To Learn The Mindset
Start thinking in systems: “When A happens, do B.”
Use tools like Zapier, n8n, or Make to create powerful automations, and your Data Engineering mindset will develop.
This is a more complex topic, which is known in the data space as “Data Products.”
Thinking in systematic ways to solve operational business challenges is an excellent method to power your transition.
Do not worry if you do not know Python right away: sketch your ideas and utilize the right no-code tools so you develop problem-solving skills. You can always learn more technical aspects later in your career!
📝 TL;DR
📊 From Dirty Looker Studio SQL to BigQuery
If you have ever used SQL to create calculated fields, you can easily get started with BigQuery.
📒 From Tracking Guide to Data Modeling
If you have ever prepared tracking guides for developers, you were already applying data modeling principles.
🦸🏻♂️ From Data Layer to APIs
Data Layer JavaScript responses are equivalent to API responses; now there is no excuse not to learn Python!
🔗 From Apps Scripts to Workflow Orchestration
Transition from scheduled Google Sheets scripts to truly powerful Python workflows.
⚒️ Use No-Code to Learn the Mindset
Use Zapier, n8n, and Make to start learning the mental models needed to make things happen.
If you enjoyed the content, hit the like ❤️ button, share, comment, repost, and all those nice things people do when like stuff these days. Glad to know you made it to this part!
Hi, I am Alejandro Aboy. I am currently working as a Data Engineer. I started in digital marketing at 19. I gained experience in website tracking, advertising, and analytics. I also founded my agency. In 2021, I found my passion for data engineering. So, I shifted my career focus, despite lacking a CS degree. I'm now pursuing this path, leveraging my diverse experience and willingness to learn.