How I Automated Weekly Analytics Reporting for an E-commerce Client

This was one of those tasks that sounds simple until you are the one doing it every week.

Open Google Analytics. Set the date range. Compare it with the same period last year. Copy the numbers somewhere. Try to explain what changed without overthinking it. Repeat next week.

No one on the team enjoyed doing it, but the insights were still important. So it kept happening, just not always consistently.

My goal was straightforward: automate the reporting process so the analysis would still exist, without someone needing to manually produce it every time.

The Real Problem Behind Weekly Reports

The client already had GA4 properly set up. Access to data was not the issue.

The problem was repetition and friction:

• Weekly comparison of the last 7 days versus the previous year
• Traffic source breakdown for e-commerce performance
• A short explanation that made the numbers readable

These steps are not complex, but they are time-consuming. Over time, they turn into background noise, and that is usually when reports start being skipped.

Building the Automation with n8n

I built a scheduled automation in n8n that runs once a week.

Each execution pulls two datasets from Google Analytics:

• Performance data from the last 7 days
• Data from the same 7-day period one year earlier

The workflow then aggregates key metrics and groups them by channel and medium. At this stage, everything is still raw but structured.

No insights, no summaries. Just clean data prepared for the next step.

Where AI Fits in This Workflow

The AI part is intentionally limited.

Instead of asking the model to analyze performance, I use it to turn structured data into readable reports. The instructions are explicit and constrained.

The model is responsible for:

• Writing a short factual summary based on the provided numbers
• Formatting tables consistently
• Adapting the output for different delivery channels

It does not invent insights or guess causes. It only translates data into text.

Delivering Reports by Email and Slack

One requirement was flexibility in delivery.

Some people prefer email. Others rely entirely on Slack. The automation supports both.

The same report is generated once and then:

• Sent as clean HTML via email
• Converted into Slack-friendly markdown for chat delivery

This avoids duplication and keeps everyone aligned around the same numbers.

What Changed After Automation

The biggest improvement was consistency.

Reports started arriving every week, at the same time, in the same format. Because they were short and predictable, people actually read them.

Questions about where the data came from decreased, since the structure never changed. Over time, the report became a shared reference rather than an extra task.

From my side, this removed hours of repetitive analysis work. For the client, it removed the mental overhead of remembering to check analytics manually.

Final Thoughts

This is not a complex analytics system, and it was never meant to be.

It is a small reporting automation that replaces a recurring manual process and fits naturally into existing workflows. In my experience, these are the automations that last.

They solve a real problem, quietly, and keep doing their job in the background.