Automating Client Reports with n8n: A Beginner's Walkthrough

Every agency and freelancer has the same Monday problem: pull last week's numbers from three dashboards, paste them into a sheet, write a short summary, send it to the client. It takes an hour per client and it is the first thing that slips when work gets busy. This walkthrough builds a simple automation in n8n that does it for you. No coding, one afternoon, and you will understand enough to automate the next boring task yourself.

What n8n is, in one paragraph

n8n is a workflow automation tool, similar in spirit to Zapier or Make, where you connect "nodes" on a canvas: a trigger (every Monday at 9 am), then steps (read a sheet, call an AI model, send an email). It has a free self-hosted version and a paid cloud version. For a beginner, the cloud trial is the quickest way to start; you can move to self-hosting later.

What we will build

1. Trigger: every Monday morning.
2. Read last week's campaign numbers from a Google Sheet (spend, clicks, leads, cost per lead, by campaign).
3. Send the numbers to an AI model with a fixed prompt that writes a 150-word client summary.
4. Email the summary to the client, with the sheet link, and send yourself a copy.

We use a Google Sheet as the data source on purpose. Connecting directly to the Google Ads or Meta APIs is possible but needs developer access and approvals; a sheet you update (or that a Looker Studio export fills) gets you the habit of automation first.

Step 1: Prepare the sheet

Create a sheet named "Weekly Report" with columns: Week start, Campaign, Spend, Clicks, Leads, CPL. Add two or three rows of last week's data and two or three for the week before. Clean column names matter; automations fail on typos.

Step 2: The trigger

In n8n, create a new workflow and add a Schedule Trigger node. Set it to run weekly, Monday, 09:00, in your time zone. Every workflow starts with a trigger; this one is the simplest kind.

Step 3: Read the sheet

Add a Google Sheets node, operation "Get many rows". Connect your Google account when prompted (n8n handles the login), choose the spreadsheet and the sheet. Run the node once with the "Test step" button; you should see your rows appear as output on the right. If you see nothing, check sharing permissions on the sheet.

Add a Filter node after it so only rows where Week start equals last Monday pass through. Expressions in n8n use JavaScript inside double curly braces; for a beginner, the expression editor's date helpers make this a two-click job.

Step 4: Turn rows into a summary with AI

Add an AI model node (n8n has nodes for the major providers; use whichever you have an API key for). In the prompt field, write a fixed instruction and insert the rows as data:

You write short weekly updates for a marketing client. Here are this week's campaign results as JSON: {{ $json }}
Write under 150 words: what happened this week in 3 bullets, the single most important change and its likely cause,
and 2 actions for next week. Plain English, no jargon, no restating every number.

Test the step. Read the output critically: if it invents a cause, tighten the prompt ("if the cause is unclear, say so"). This is the same discipline as writing prompts by hand; see our guide to ChatGPT prompts for marketers.

Step 5: Send it

Add a Gmail (or SMTP) node. Set the recipient, a subject such as "Weekly campaign update, week of {{ date }}", and paste the AI output into the body with the sheet link below it. Add a second email node, or a CC, so you get a copy and can catch mistakes before the client replies.

Prefer WhatsApp? n8n has HTTP and WhatsApp Business nodes, but the Business API needs approval, so start with email and add WhatsApp once the report is trusted.

Step 6: Turn it on and watch it twice

Activate the workflow. For the first two weeks, read every report before it reaches the client; keep a manual version in parallel. Automations fail quietly when a column is renamed or a login expires, so n8n's "Error Trigger" node, which messages you when something breaks, is worth adding on day one.

Where to go next

• Replace the manual sheet with a scheduled export from Looker Studio or the ad platform.
• Add a second branch that posts the summary to a Slack or Teams channel for your own team.
• Build a lead-routing workflow: form submission in, enrichment and a WhatsApp follow-up out.

Automation and AI tooling are taught as part of the Digital Marketing Professional Program, where students build workflows like this one for live projects. Browse the full list of digital marketing courses in Rajkot to see which track fits you.