Weekly reports are the tax you pay for having a team. They’re necessary, but the process is almost always manual: pull data from three tools, paste it into a spreadsheet, format it, email it, and hope you didn’t fat-finger a number. I’ve watched founders spend two hours every Friday on this. That’s over 100 hours a year.
You don’t need a developer to fix it. You need a script that runs on a schedule. Here’s how to build one using tools you already have.
First, map the data sources. Most weekly reports pull from a CRM, a project tracker, and maybe a billing system. Write down exactly which fields you need: new leads, closed deals, tasks completed, overdue invoices. Don’t worry about formatting yet. Just list the raw inputs.
Next, check if each tool has an API. Most modern SaaS products do. If yours doesn’t, look for a CSV export URL. Some tools let you generate a one-click export link that updates daily. That’s often enough. If you’re stuck, a tool like Zapier or Make can bridge the gap, but I’d avoid them for this if you can. They add cost and another failure point.
Now pick a scripting language. Python is the default for a reason: it’s readable, has libraries for everything, and runs anywhere. You don’t need to be a Python expert. If you can write a formula in Excel, you can write this script. The core logic is: fetch data, transform it, write it to a file, send it.
Start with one data source. Write a script that pulls the data and prints it to the console. Don’t worry about the rest yet. Once that works, add the second source. Then the third. You’re building a pipeline, and each piece should be testable on its own.
For the transformation step, keep it simple. Convert dates to a consistent format. Sum the numbers you care about. If you need a table, use a library like Pandas or just build a CSV. The goal is a clean, predictable output, not a work of art.
Formatting is where most people overthink. You don’t need a beautiful HTML email. A plain-text summary with a few bullet points is fine. If you want something nicer, generate a simple HTML table. But the value is in the data, not the design.
Scheduling is the final piece. On macOS or Linux, cron is built in. On Windows, Task Scheduler works. If you want zero maintenance, deploy the script to a serverless function (AWS Lambda, Google Cloud Functions, or a cheap VPS). Set it to run every Friday at 8 AM. Then forget it.
One tip: log everything. When the script fails, you want to know why. Write a log file with timestamps and error messages. It’ll save you hours of debugging later.
If you’d rather not build from scratch, our Weekly Report Automator does all of this out of the box. It connects to 15 common tools, includes a template for the email, and runs on a schedule. You just enter your API keys. Either way, the principle is the same: automate the boring part, and spend your Friday on something that matters.